Week of November 10, 2025
Rhinestone letters spelling "Compton" caught my eye as I slid into an Uber heading to the Nashville airport last week. The black baseball cap sat prominently on the dashboard—an obvious conversation piece that led to an unexpectedly confessional discussion about my driver's time in the Crips and his journey out of the LA gang, despite pressure from his family and friends.
I'm continually amazed by how many drivers embrace the "share" part of "rideshare." (And, I still want a button that lets drivers tip riders for being good listeners—or for facilitating meaningful emotional breakthroughs.)
I deserved a tax-free 20% tip for this ride. My driver complimented the depth and nuance of my questions. Caring for dialysis patients involved in Baltimore's drug trade provided me a surprisingly functional grasp of the day-to-day HR, logistics, and political realities of gang life.
---
Listen to a Google Notebook LM A.I.-generated podcast of the newsletter with two virtual "hosts."
https://drive.google.com/file/d/1bwCwUa6HfghTNXj3q6UZISXNFXGdrJfF/view
------
Science and Technology Trends
Data on Smartwatches! I found a meta-analysis evaluating the diagnostic accuracy of commercially available smartwatches in detecting atrial fibrillation. The analysis included 26 studies and 17,349 patients. In general, all smartwatche brands demonstrated excellent diagnostic performance with a pooled sensitivity of 95% and a specificity of 97% (across all the studies) . However, real-world applicability (specifically regarding the value of screening large populations, many of whom are low-risk, and the impact of false-positive results) makes scaled clinical implementation less compelling for this use case.
Article: https://www.jacc.org/doi/epdf/10.1016/j.jacadv.2025.102133
AI Summary: https://claude.ai/public/artifacts/eac2c3ad-487d-48a7-bf0d-7c770052cf57
More clinical data on coffee, which (I increasingly believe) may be the most amazing neuro-enhancing substance on earth - minimal downside, legal, and efficacious. In this case, researchers from Inner Mongolia Medical University (yes, I know it's not a typical research institution) analyzed NHANES data (a large, multi-year dataset from the U.S.) to examine coffee/caffeine consumption, as well as cognitive performance, particularly in memory and processing speed, among older adults. The analysis found a strong, positive relationship between self-reported high-volume coffee intake (≥480 g/day) and 42% lower odds of poor cognitive performance. Decaffinated coffee did not offer the same benefit. The authors were examining potential biochemical mechanisms (with considerable discussion on specific liver enzyme tests), but I was most excited by the ongoing safety and efficacy data.
Article:
https://nutritionj.biomedcentral.com/counter/pdf/10.1186/s12937-025-01173-x.pdf
AI Summary
https://claude.ai/share/d00d7bfd-c841-4b56-989a-9ec07848e3a8
Japanese Scientists published this article about ways to interpret fMRI data - translating brain blood flow data into written descriptions of videos that human subjects were watching or imagining. The AI achieved 50% accuracy when choosing from a limited library of 100 different videos. The fMRI interpretation was done without focusing on the brain's typical language areas, suggesting that visual thoughts have their own structure and patterns. This technology may one day help people who can't speak due to conditions like stroke or ALS; it requires many hours of practice and data. Additionally, the technique raises concerns explored in many science fiction tropes about privacy and unwanted access to our thoughts.
Article
https://www.science.org/doi/epdf/10.1126/sciadv.adw1464
Summary
https://claude.ai/public/artifacts/7d169f85-e319-4f29-9bd2-ab1a7e4d28ed
Anti-Science Articles of Note
Here is your weekly reminder: correlation ≠ does not imply causation.
https://x.com/mrnllmtt/status/1985356887173501235?s=42
Eric Topol interviewed Epidemiologist Seth Berkley on the global infrastructure, distribution, science, and politics of vaccines. The discussion is fascinating and a good example of how one can work with the best available science to make big bets, even in the face of uncertainty.
https://open.substack.com/pub/erictopol/p/seth-berkley-and-eric-topol-discuss?r=eywiv&utm_medium=ios
Related -
This article, published in July 2025, was featured in various news outlets and reposted on social media last week. "Evaluation of the Effects of mRNA-COVID 19 Vaccines on Corneal Endothelium" is a small study out of Turkey examining 128 eyes from 64 patients to assess corneal endothelial changes approximately 2.5 months after receiving the Pfizer-BioNTech mRNA COVID-19 vaccine. The study found statistically significant changes, including a decrease in endothelial cell density (ECD), an increase in central corneal thickness (CCT), a reduction in hexagonality, and an increase in the coefficient of variation. No patients reported vision changes. And, even though this is a small, highly imperfect prospective cohort review, social media has embraced these data as another example of the "conspiratorial harm being kept from the sheeple" who willingly accept "corrupt science" claims. Sadly, the social media posts seem to miss the more critical analysis of the research including: no control group, a small sample size (64 patients), single vaccine type studied, short follow-up period (2.5 months), no baseline test-retest variability data, potential selection bias, no adjustment for confounders (age, comorbidities, medications, COVID infection history), lack of biological mechanism explanation, unclear and clinical significance of findings. My interpretation: "Vaccines have a lot of compelling data with scaled, measured value. Nevertheless, this study finds some new, interesting data that requires way more research, and these observations may give us some hypotheses to explore. And, I would still recommend getting the COVID vaccines."
Article:
https://pubmed.ncbi.nlm.nih.gov/40658089/
Typical Internet Hyperbole:
https://www.facebook.com/news.now.go/posts/new-study-finds-pfizer-covid-vaccine-may-cause-serious-eye-damage-is-your-vision/1212205094343611/
and
X debates (an endless text stream of people screaming past each other):
https://x.com/ejc3/status/1985596908409438348
Living with AI.
This New Yorker article is worth reading with coffee. The author believes humans suffer from "substrate chauvinism" - unfairly dismissing AI intelligence because it doesn't use biological neurons like human brains. He argues that AI systems achieve genuine understanding through different mechanisms (vector mathematics, compression) but produce functionally similar cognitive outcomes to humans - recognizing patterns, making connections, and solving problems. Drawing on various theories of human cognition ("seeing as" theory and compression-understanding connection), it appears that AI is a form of silicon-based intelligence that we fail to recognize as "real thinking" simply because it's alien to our carbon-based cognition. We may have created a new form of intelligent life that thinks differently but genuinely, and our biological bias prevents us from appreciating this achievement.
https://www.newyorker.com/magazine/2025/11/10/the-case-that-ai-is-thinking
AI Summary: https://claude.ai/public/artifacts/451ef2d4-80ea-4d16-a379-b2877ea6d567
On the other hand, this Nature article highlights the many failings of ChatGPT as a healthcare consultant. "Version 5 of the LLM is a probabilistic text generator that can confidently deliver incorrect medical information, break safety rules, and retain latent biosecurity capabilities. The improved fluency creates an "illusion of understanding" that may be more dangerous than outright errors in clinical settings, where GPT-5 still fails more than half of difficult safety-critical scenarios despite recent improvements."
Article: https://www.nature.com/articles/s41591-025-04008-8
AI Summary of the AI article:
https://claude.ai/public/artifacts/928f2cec-51c5-40cb-b835-c29642980496
Related commentary from STAT News: "Doctors need to ask patients about chatbots; It should be a standard part of the intake process across specialties."
https://smry.ai/https://www.statnews.com/2025/10/29/chatbots-doctors-guide-medical-appointments-questions/
Infographics
A fantastic, very busy one-page explainer on hemodialysis in the ICU.
https://static1.squarespace.com/static/5e6d5df1ff954d5b7b139463/t/602df4ca0182c74344062da1/1613624529281/ICU_one_pager_RRT.png
from
https://onepagericu.com/index
Things I learned this week
I learned about the phrase "Arm Farms," referring to businesses in South Asia where young engineers strap GoPros to their foreheads and perform tasks such as folding laundry or packing boxes to teach humanoid robots how to do chores.
https://www.latimes.com/business/story/2025-11-02/inside-californias-rush-to-gather-human-data-for-building-humanoid-robots
I learned that NPR has a section called "Goats and Soda - Stories of Life in a Changing World." In this section, I learned about the newly captured footage of a rat capturing and eating a bat. Aside from the odd nature of this story, there are several takeaway points of note:
1) There is a journal article on this topic.
https://www.sciencedirect.com/science/article/pii/S2351989425004950?via%3Dihub
2) Rats biting bats is a possible way for rats to get all sorts of unpleasant bacteria or viruses (which could then spread elsewhere).
3)Do not mess with Northern German Rats (Rattus norvegicus). I have deep respect for an animal that can snatch a flying bat out of the air mid-flight.
https://www.npr.org/sections/goats-and-soda/2025/11/08/g-s1-96931/rat-bat-virus-video
AI art of the week
A visual mashup of topics from the newsletter.
I use ChatGPT to summarize the newsletter, suggest prompts, and generate images with several LLMs.
A 1960s surreal advertising illustration blending gritty human storytelling with industrial optimism. A rhinestone-studded black baseball cap reading "Compton" rests on the dashboard of a stylized taxi, whose interior transforms into a laboratory assembly line. Inside, young engineers with GoPros on their foreheads calmly teach humanoid robots to fold laundry and make coffee. At the same time, a thoughtful driver glances into the rearview mirror—his reflection composed of halftone dots and chrome geometry. In the background, a heroic mid-leap rat catches a bat, drawn with vintage detergent-ad dramatization. Overhead text in retro slab serif reads "Confessions from the Arm Farm — Emotional Labor, Automated."
Art Direction & Palette:
Aesthetic: obscure 1960s industrial ad style—mix of Erik Nitsche's scientific minimalism and Fredun Shapur's abstract playfulness, with subtle Push Pin Studios collage humor.
Color palette: mustard yellow, avocado green, slate blue, coral pink, chrome silver, faded black ink.
Composition: geometric layout divided into modular "panels," reminiscent of mid-century product diagrams and atomic-era promotional posters.
Texture & Typography: halftone gradients, matte grain, and letterpress slogans in mixed Futura Condensed and Clarendon.
Mood: hopeful, confessional, slightly uncanny—where human empathy meets mechanized imitation under the sheen of mid-century commercial optimism.
More specific prompts yield greater consistency, but homonyms seemed to trip up Grok (look at the bat).
Gemini
https://drive.google.com/file/d/1aymdSgcTVYm9AMCcAXmcPN5qGHM13tmm/view
Grok
https://drive.google.com/file/d/1aymdSgcTVYm9AMCcAXmcPN5qGHM13tmm/view
Perplexity
https://drive.google.com/file/d/1prpNe6nBXcFNzRrDSjaOfZCgFQ8ho4ss/view
ChatGPT
https://drive.google.com/file/d/15mUcjI_8bC2JMnOaCPoAI6T5-NEYpt_1/view
---
The Pandemic Mitigation Collaborative (PMC) utilizes wastewater viral RNA levels to forecast four-week predictions of COVID-19 rates.
CDC and other data are still offline.
We are in a lull, however. A good time to get your vaccines before the holidays.
https://pmc19.com/data/
based upon https://biobot.io/data/
Wastewater Scan offers a multi-organism wastewater dashboard with an excellent visual display of individual treatment plant-level data.
https://data.wastewaterscan.org/
----
Clean hands and sharp minds,
Adam
Relive all the thrills and excitement - The What Adam is Reading Archive
http://www.whatadamisreading.com/
CONFIDENTIALITY NOTICE: THIS MESSAGE IS CONFIDENTIAL, INTENDED FOR THE NAMED RECIPIENT(S) AND MAY CONTAIN INFORMATION THAT IS (I) PROPRIETARY TO THE SENDER, AND/OR, (II) PRIVILEGED, CONFIDENTIAL, AND/OR OTHERWISE EXEMPT FROM DISCLOSURE UNDER APPLICABLE STATE AND FEDERAL LAW, INCLUDING, BUT NOT LIMITED TO, PRIVACY STANDARDS IMPOSED PURSUANT TO THE FEDERAL HEALTH INSURANCE PORTABILITY AND ACCOUNTABILITY ACT OF 1996 ("HIPAA"). IF YOU ARE NOT THE INTENDED RECIPIENT, OR THE EMPLOYEE OR AGENT RESPONSIBLE FOR DELIVERING THE MESSAGE TO THE INTENDED RECIPIENT, YOU ARE HEREBY NOTIFIED THAT ANY DISSEMINATION, DISTRIBUTION OR COPYING OF THIS COMMUNICATION IS STRICTLY PROHIBITED. IF YOU HAVE RECEIVED THIS TRANSMISSION IN ERROR, PLEASE (I) NOTIFY US IMMEDIATELY BY REPLY E-MAIL OR BY TELEPHONE AT (855.472.9822), (II) REMOVE IT FROM YOUR SYSTEM, AND (III) DESTROY THE ORIGINAL TRANSMISSION AND ITS ATTACHMENTS WITHOUT READING OR SAVING THEM. THANK YOU.
-DaVita Inc-
Comments
Post a Comment