Week of July 7, 2025
I have come to enjoy Fourth of July parades. Politicians, small businesses, and random organizational floats offer a fascinating glimpse into a community. The highlight of this year's Severna Park, Maryland, parade was the motorized, remote-controlled spot-a-potties. Gotügo (a local portable restroom company that lives up to the "projected sophistication" implied by their name's inclusion of an umlaut) shared this innovative approach to parading toilets. I now see a future filled with mechanized, autonomous AI-driven toilets (which feels slightly less intimidating than Boston Dynamics' robot dogs).
See the mechanized toilets!
https://share.icloud.com/photos/042gwn0WwE5ozuPMEHV5m8tZA
---
Listen to a Google Notebook LM A.I.-generated podcast of the newsletter with two virtual "hosts," one of whom (this week) believes she wrote the newsletter.
https://drive.google.com/file/d/1_IK_GrnhE4Px5pwvAxYiuBj0hH-xOlA4/view
------
Science and Technology Trends
A group of UK researchers published data on how human gut bacteria can effectively metabolize per- and polyfluoroalkyl substances, also known as "forever chemicals" (PFAS). In this study, 38 out of 89 types of tested human gut bacteria actively concentrate PFAS intracellularly within minutes of exposure. The study was limited to mice (though the gut bacteria were known human species) and only a single PFAS exposure. However, these data suggest gut microbiota (the human microbiome) may help mitigate PFAS exposure.
Article: https://www.nature.com/articles/s41564-025-02032-5
and
AI Summary: https://claude.ai/public/artifacts/b7f3a73f-d7a7-4da8-bfd8-f98aab3c0f0e
French researchers recently identified a 48th human blood group type. Most of us are familiar with the A/B/AB/O and Rh+/Rh- designations. These are critical for ensuring transfusion safety. However, 46 other known groupings of red cell membrane proteins (less critical in transfusions) play a role in immune functions (amongst other things). Each human has proteins from all 48 groupings (or systems) on their red blood cells. A "complete" blood type would describe the phenotype for all 48 systems, not just ABO and Rh. Regardless of the details, this 48th group resulted from a genetic mutation in one person from the island of Guadeloupe. It serves as a snapshot of how evolution works (either it dies out with this person, passes on slowly, or if the protein confers some advantage, the mutation may appear more commonly in future generations).
https://www.popularmechanics.com/science/health/a65161270/gwada-negative-48-blood-group/
and
AI summary:
https://claude.ai/public/artifacts/2a8d74e3-bb19-4934-b720-3b22b7c4e4b3
Over the holiday weekend, after discussing cats with visiting family, several feline-oriented posts appeared in my social media feed (note that we do not own cats). One post referenced a recent meta-analysis but only had the headline "cat ownership is associated with 2x odds of schizophrenia." The family member (who has a PhD in neuroscience) and I laughed at the implied causation. However, I was bothered by the post with a provocative headline that may fallaciously imply the statement is a known, settled truth. I decided to explore the data further:
Here is the referenced journal article: https://pubmed.ncbi.nlm.nih.gov/38041862/
The authors' conclusions are far from definitive: "Our findings provide support for the hypothesis that cat exposure is associated with an increased risk of broadly defined schizophrenia-related disorders; however, these are mixed findings. There is a need for more high-quality studies in this field."
I had Google Gemini generate a comprehensive research summary of the available evidence linking schizophrenia to cat ownership by compiling a summary of studies and evaluating the data. Here is the 19-page(!) deep research report:
https://docs.google.com/document/d/1HvC4gx6_cujtZFDS4w1pFEJLv0MwMpFbVI_FImBbd8A/edit
The TL;DR - Gemini's noteworthy summary of 42 identified resources:
- There is no definitive causal link between general cat ownership and the development of schizophrenia. The observed association is more likely the result of a complex interplay of multiple factors, including:
- A potential true, but likely small, causal effect from specific high-risk exposures (e.g., owning a rodent-hunting cat, experiencing a cat bite) in a subset of genetically vulnerable individuals.
- A statistical artifact created by uncontrolled confounding variables, such as socioeconomic status, household chaos, or parental mental health.
- An artifact of recall and selection biases inherent in the low-quality, retrospective case-control studies that form the bulk of the positive evidence.
- A degree of reverse causation, where the early stages of mental illness may lead an individual to seek feline companionship.
The [unsurprising] real conclusion is that it is far easier to generate partially false information that seems true than to spend time and energy disproving such statements.
Anti-Anti-Science Articles of Note
More (reasonable and harsh) comments on the CDC's vaccine advisory board meeting, ACIP.
https://www.medpagetoday.com/infectiousdisease/vaccines/116343
I have mixed feelings over our HHS secretary's recently stated goal of "all Americans using wearable tech" to improve their health. On the one hand, focusing on movement and preventative health seems reasonable. On the other hand, wearables come with significant costs, questionable value, and such a program would occupy the public's finite "attention bandwidth." ABC News offered a reasonable article highlighting the tradeoffs of focusing attention on the topic of wearables:
https://abcnews.go.com/Health/rfk-jr-wearables-benefits-risks/story
and
I had Google Gemini generate a comprehensive report on the available evidence, examining wearables as a cost-effective means of improving health outcomes. Dozens of studies (of varying quality) examine a wide range of disease states and health outcomes. The bottom line is that wearables have some value in narrow populations, but it is unclear what problems they will solve in a large, heterogeneous population.
https://docs.google.com/document/d/1RrYMjAtHuf0JpEcjT2HZ-OVkoslPj3mwnRQ06vuZxG8/edit
Here is a very summarized version of the conclusions by disease state:
- Obesity / Weight Management: Generally Ineffective or Counterproductive for Long-Term Weight Loss.
- Cardiometabolic Conditions: Effective for Increasing Activity, but Limited Impact on Clinical Markers.
- Atrial Fibrillation (AFib): High Diagnostic Potential for Screening, but with Significant Downsides.
- Hospitalized Patients: Effective for Combating Deconditioning.
- Rehabilitation & Functional Improvement: Highly Effective in Niche Clinical Applications.
- Infectious Disease Detection: Promising but Not Yet a Validated Clinical Tool.
- Fall Detection: Limited Clinical Utility Due to High False Alarm Rate.
- Health Anxiety & Psychological Well-being: Negative Psychological Impact in Vulnerable Individuals.
Living with AI.
The New Yorker published a well-written editorial on the rise and ramifications of AI use amongst college students. The author captures the tradeoffs of using AI to work hard vs. work smart, all while still learning something.
https://www.newyorker.com/magazine/2025/07/07/the-end-of-the-english-paper
Marc Andreessen's 2025 career advice to working in an age of AI:
https://x.com/benln/status/1940106102668210273/photo/1
I tried something new this week. I found Stanford health policy researcher, Dr. Sherri Rose's JAMA editorial on the hype vs. reality of using AI in healthcare:
https://jamanetwork.com/journals/jama-health-forum/fullarticle/2835107
I pulled the JAMA article's references and added a few other recent healthcare AI articles into Notebook LM to create a briefing document:
https://docs.google.com/document/d/1Y-pzrGyBhR7jQVllCYWcunk3VNSESy90vMGLn9ma5kg/edit?usp=sharing
I then asked Claude to compare the Notebook LM-generated briefing document with the Rose article, both of which used a similar set of articles.
https://claude.ai/public/artifacts/08056559-775c-439d-887b-b21677efabc1
Conclusion from Claude:
- "Both documents represent different but complementary perspectives on AI in healthcare:
- Rose provides the "regulatory voice" - emphasizing caution, proper evaluation, and the dangers of premature implementation - essentially preventing harm and ensuring responsible development.
- The Briefing Document provides the "technical roadmap," outlining specific methodologies, performance standards, and implementation strategies needed to realize AI's potential safely and effectively.
- Together, they paint a picture that AI in healthcare has transformative potential, but realizing this potential requires rigorous evaluation processes. The key insight from both pieces is that the problem isn't AI itself, but rather the current approach to its development and deployment in healthcare settings. Both argue for a more measured, evidence-based approach that prioritizes patient safety and clinical utility over technological novelty."
Infographics
I have never taken the time to examine the composition of different types of magnets. There are four types, all based on iron, cobalt, or nickel.
https://cen.acs.org/materials/Periodic-Graphics-Permanent-magnets-explained/103/web/2025/04
Things I learned this week
My favorite "blamouflage" headline of the week: "After 41 years in captivity, this celebrity sea turtle is finally swimming home. Here's how Jorge the sea turtle prepared for the improbable journey—decades after fishermen found him tangled in a fishing net off the coast of Argentina." On first read, it sounds like Jorge the turtle took a 4-decade voluntary, relaxing respite in an Argentinian aquarium and (recently) decided to go home. Regardless of Jorge's agency (or lack thereof), it appears that his Mendozian caretakers developed his skill set and prepared him for reintroduction to the wild.
Paywall
https://www.nationalgeographic.com/animals/article/sea-turtle-rehabilitation-return-to-ocean
No Paywall
https://www.smry.ai/proxy?url=https%3A%2F%2Fwww.nationalgeographic.com%2Fanimals%2Farticle%2Fsea-turtle-rehabilitation-return-to-ocean
An Annapolis (Maryland) bagel shop offered red, white, and blue bagels, along with pop-rock-infused berry cream cheese, in celebration of the Fourth of July. The strange sensation of cream cheese with Pop Rocks made me realize I didn't know what was exploding in my mouth or why. I now understand the history and chemistry behind one of the more unique candies from my childhood.
https://redstonefoods.com/candy-soda-blog/history-of-popping-candy
The article was particularly evocative - reminding me of a childhood urban legend (told at summer camp as I heard it) that the Life Cereal kid, Mikey, from the "Mikey likes it!" commercials had tragically died from eating Pop Rocks and drinking a Coke, "causing his stomach exploded from all the excess gas." Mikey, also known as John Gilchrest, is alive, and to the best of my knowledge, has a fully intact stomach.
https://en.wikipedia.org/wiki/Little_Mikey
and
https://medium.com/the-daily-cuppa-grande/what-ever-happened-to-mikey-59d76992c49e
A loyal reader introduced me to organized "professional" pillow fighting.
https://fightpfc.com/
It appears that any activity can be turned into a sport with a good promoter and sufficient funding. Indeed, between the discovery of the best "combat pillows" and recruiting existing athletes from boxing, wrestling, and MMA, the founders of Pillow Fight Championship (PFC) have created an entertaining sport that offers "a significant cardiovascular workout."
https://www.forbes.com/sites/candaceoehler/2024/11/28/dont-sleep-on-professional-pillow-fightingits-real/
AI art of the week
A visual mashup of topics from the newsletter.
I use ChatGPT to summarize the newsletter, suggest prompts, and make the images.
https://drive.google.com/file/d/11f1B7Kx82oYKE3_Pjb5aRgBaGin0QqKg/view
"Paint a dignified Fourth of July parade in a small American town, using the meticulous, nostalgic realism of Norman Rockwell. The scene shows two autonomous, AI-powered portable toilets rolling solemnly down the main street, adorned with patriotic bunting and antennas. A float follows, carrying heroic, anthropomorphized bacteria standing atop a giant human intestine, holding glowing spheres representing captured PFAS chemicals. Onlookers—including children and adults—watch with composed, thoughtful expressions. In the background, a 'Pillow Fighting Championship' unfolds on a raised platform, with soft bursts of Pop Rocks-style sparks. The crowd remains engaged but stately, evoking mid-century Americana with subtle surrealism and rich emotional undertones."
---
The Pandemic Mitigation Collaborative (PMC) utilizes wastewater viral RNA levels to forecast four-week predictions of COVID-19 rates.
Wastewater RNA levels predicted a COVID prevalence rate from 1 in 167 (week of June 23) to 1 in 197 last week (week of June 30).
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
Comments
Post a Comment