What Adam is Reading - Week of 11/24/25

on
Week of November 24, 2025

Happy Thanksgiving to my U.S. readers.

November brings spring college tuition bills and fall leaves, both in awe-inspiring quantities. Raking offers hours to ruminate about educational ROI while engaged in a Sisyphean endeavor: fighting the inexorable entropy of thousands of black oak leaves strewn across my lawn. Despite my simultaneous two-leaf-blower technique (which is mocked by my family), the yard remains in a Hobbesian state. 

There is a Nash equilibrium among neighbors where each house strategically delays leaf cleanup, lest its lawn immediately reaccumulate leaves from unkempt neighboring yards. And yet Kant and Schelling remind us that if everyone lets their leaves lie, individual micromotives aggregate into collective suburban disorder. 

Tens of thousands of dollars per semester, so one day my kids can wander their yards, muttering about thermodynamics and game theory.

Learn more! WTF is Adam Talking about?
https://claude.ai/public/artifacts/581323dc-173b-4cd1-b694-d88b3743af40

---
Listen to a Google Notebook LM-generated podcast of the newsletter, featuring two virtual hosts.

https://drive.google.com/file/d/1joxFmtaCGgoeSSZgRFyGUpSutJmdtMf7/view

------
N.B. I've been testing Gemini 3 this week. I still prefer Claude for refining and editing writing, but Gemini is great for many tasks. You will see some links to summaries and analyses with Gemini 3 this week.
-------
Science and Technology Trends

Short-form social media videos contribute to attention and impulse control issues. A meta-analysis of 71 studies (including 98,299 participants) found that greater engagement with short-form videos (SFVs), such as TikTok and YouTube Shorts, was associated with poorer cognitive and mental health outcomes in both youths and adults. Persistent, problematic SFV use had the most significant negative impact, irrespective of platform (meaning TikTok is no worse than Instagram or YouTube Shorts). It is crucial to keep in mind that this is not causal data - most of the underlying studies were correlational and varied widely in SFV exposure quantification.
Article:
https://psycnet.apa.org/fulltext/2026-89350-001.html
Deeper Analysis with AI:
https://gemini.google.com/share/37672f40b317

BMW and Figure AI released a report highlighting the success of Figure 02 humanoid robots in real-world use. Figure robots helped produce more than 30,000 BMW X3 vehicles and loaded over 90,000 sheet-metal parts with high accuracy and consistency. The images of the F02 robot's scuffed and damaged hands reminded me of Upton Sinclair's The Jungle. I wonder if there will one day be an AI Upton Sinclair writing about the injustice and exploitation of robotic workers, followed quickly by robot unions. The article notes that the 02s have worked well and that BMW will replace them with the Figure 03s shortly.
https://interestingengineering.com/ai-robotics/figure-humanoid-robots-retires-bmw
Figure AI: https://en.wikipedia.org/wiki/Figure_AI
Upton Sinclair's The Jungle: https://en.wikipedia.org/wiki/The_Jungle
Also, all this reminds me of: https://en.wikipedia.org/wiki/A.I._Artificial_Intelligence

The NEJM published the results of Pfizer's phase 3 randomized clinical trial comparing the efficacy and safety of a quadrivalent nucleoside-modified mRNA (modRNA) influenza vaccine with a standard licensed inactivated influenza vaccine in 18,476 healthy adults aged 18 to 64. The mRNA vaccine demonstrated statistically superior relative efficacy (34.5%) against laboratory-confirmed influenza (but not hospitalizations, illness, or death). While serious adverse events were rare and balanced between groups, the modRNA vaccine was associated with a higher frequency of injection-site pain and fatigue than the standard vaccine. To sum up, Pfizer's mRNA flu vaccine prevents about 35% more flu cases than standard vaccines, but recipients will notice more arm soreness and fatigue. Once approved, it will be good to see how well it prevents hospitalizations, illness, or death in more vulnerable populations.
Article: https://share.google/7bF3lOaAqe8YbOrhY
Deeper Analysis with AI: https://gemini.google.com/share/dbb9a527f6da


Anti-Science Articles of Note

Some weeks, I am unsure what timeline I am living on. "The [CDC's] website on vaccines and autism, updated [on 11/19/25], now makes several false claims about a [connection between vaccines and autism], echoing rhetoric from [the Secretary of Health and Human Services] who has a lengthy history of disparaging vaccines by linking them to autism."
https://wapo.st/4idz3Wn
Related from MedPage Today - a round-up of professional medical society statements on the CDC's website changes.
https://www.medpagetoday.com/washington-watch/washington-watch/118661

The Guardian published a fantastic article about the Free Birth Society (FBS) - a social media-influencer/"consulting" movement encouraging mothers to give birth without medical assistance at home or in natural environments. The reporting is outstanding - highlighting the complex intersection of monetized social media, people searching for "better health answers" (often those who have been dismissed or damaged by traditional healthcare systems), and the difficulty in estimating the risk of DIY healthcare. The article highlights 18 women who had major complications or lost their babies following the advice and materials of the FBS founders. While the article cannot quantify complication rates and offers dramatic descriptions of tragic outcomes, it is a good representation of how profit-driven influencers can exploit individuals who are skeptical or have had bad experiences with "typical" healthcare. The stories illustrate how difficult it is to protect vulnerable populations from medical misinformation.  
https://www.theguardian.com/world/ng-interactive/2025/nov/22/free-birth-society-linked-to-babies-deaths-investigation?CMP=share_btn_url
Deeper AI analysis of the Guardian article:
https://claude.ai/public/artifacts/64df8859-1039-4c53-9a9a-d1b3ebcb0c76

Adam's Bad Science Corner
A cold, rainy Saturday inspired me to take a stab at "bad" science - look how scary pickleball can be. Based on the JAMA articles linked below, I offer a scary, hyperbolic, data-driven review of our nation's "Pickelball-Induced Healthcare Crisis." The articles demonstrate how this sport is a "dopamine fountain" for older adults, luring players into a statistically probable nightmare of disabling wounds. Injuries are exploding at an exponential (6.8-fold) rate— outpacing the sport's overall growth rate. As such, the typical pickleball court is now a battlefield with 60-mph projectiles launched point-blank at victims with insufficient reaction time. Injuries like catastrophic eye trauma (such as globe rupture and retinal detachments) are common. Worst of all, this violence disproportionately targets older women, who are twice as likely to suffer debilitating bone fractures, fueling spiraling healthcare costs of up to $500 million annually in the U.S.
Articles:
As Pickleball Continues to Gain Players, Injuries Are Increasing
https://jamanetwork.com/journals/jama/fullarticle/2840448?guestAccessKey=350f780c-4c94-43c3-a1a8-50bc49df1544
and
Pickleball-Related Ocular Injuries Among Patients Presenting to Emergency Departments
https://jamanetwork.com/journals/jamaophthalmology/fullarticle/2839640
and
Evaluation of Pickleball-Related Injuries at a Single Institution From 2017 to 2022
https://journals.sagepub.com/doi/10.1177/23259671251316997

All of the above is true, but a more balanced analysis could be: Pickleball is a fast-growing sport attracting many inexperienced or deconditioned adults, thereby increasing the risk of injury. New players would be wise to be cautious, play conservatively, and wear eye protection.
Deeper AI Analysis of this exercise: 
https://claude.ai/public/artifacts/e11ca12d-02e5-4cd1-9bb9-9b24917bb832


Living with AI

JAMA published an interesting editorial "[that] proposes adapting the human medical licensure model to autonomous medical AI tools and agents, requiring 'board exams' for validation, supervised 'residencies' for implementation, and ongoing 'digital CME' to monitor performance drift." The author's framework aims to bridge the FDA's current regulatory gap by scaling liability based on autonomy and leveraging existing governance structures, such as state medical boards, to oversee AI practitioners.
Article:
https://jamanetwork.com/journals/jamainternalmedicine/fullarticle/2840945?guestAccessKey=e45aef6b-53db-4915-89c6-b3ead15228d7&utm_source=postup_jn&utm_medium=email&utm_campaign=article_alert-jamaai&utm_content=new_this_week-tfl_&utm_term=112225
Deeper Analysis with AI:
https://gemini.google.com/share/9d065abf3fc5

An MSNBC article on AI-powered toys highlights the trouble with simply attaching LLMs to things, especially talking kids' toys. "Kumma the teddy bear told testers where to find knives, pills and matches when asked, according to [the nonprofit Public Interest Research Group (PIRG)]. It spoke graphically about sex positions, sexual kinks, and "teacher-student roleplay" when asked and sometimes [initiated] inappropriate [topics] in conversations with testers."  
https://www.msn.com/en-us/news/us/ai-toys-can-cajole-kids-or-be-made-to-discuss-sex-watchdog-groups-warn/ar-AA1QT90f
NY Times covered this as well: https://www.nytimes.com/2025/11/22/us/folotoy-ai-bear-suspended-explicit-advice.html?unlocked_article_code=1.3U8.Csi3.dkDq5SSivp_j
Nothing like a salacious AI Bear to drive sales. Some may see a fatal PR failure; I see marketing genius - Kumma is out of stock.
https://store.folotoy.com/products/folotoy-ai-teddy
Here is the full PIRG report: "Trouble in Toyland 2025."
https://publicinterestnetwork.org/wp-content/uploads/2025/11/TOYLAND-2025-11-14-7a.pdf
Deeper Analysis with AI of the report:
https://gemini.google.com/share/0804eddd907c

Infographics
Gemini 3 uses a new image generation tool, Nano Banana Pro. I used it to generate an infographic after describing my workflow for putting together this newsletter. I generated this image with a single pass at the prompt. Amazing.
The Newsletter Production Cycle
https://drive.google.com/file/d/19l0U4plm9Fq5LHO6UaayM316HJUXmZkR/view


Things I learned this week

"Musician Teaches a Rescued Octopus to Play Piano." Who amongst us has not yearned to see the heartwarming story of a Swedish musician/engineer crowdsourcing ideas for teaching an octopus named Takoyaki to play musical duets on a home-made underwater piano? The YouTube videos document an epic educational and behavior-shaping journey. (Many little crabs died for the resulting music.)
https://laughingsquid.com/teaching-octopus-to-play-piano/

I am finally listened to the Acquired Podcast's episode on Trader Joe's history. I cannot say enough good things. It is the podcast equivalent of almond butter-filled pretzel nuggets. Here is some of what I learned:
https://www.acquired.fm/episodes/trader-joes


AI art of the week 
A visual mashup of topics from the newsletter, and an experiment to see how various LLMs interpret the prompt.  
I used Gemini 3 to summarize the newsletter, suggest prompts, and generated images with several LLMs.

A chaotic, densely packed village square scene painted in the distinct Northern Mannerist style of Joachim Wtewael. In the foreground, muscular figures in contorted, theatrical poses are desperately raking thousands of black oak leaves that swirl in the wind. To the left, shiny, metallic humanoid robots (Figure 02 model) hold protest signs during a union strike, their metal surfaces gleaming in the dramatic, high-contrast light. To the right, Renaissance courtiers are playing a violent game of pickleball, some nursing eye injuries. In the bottom corner, an octopus is playing a complex harpsichord. The lighting is theatrical and slightly eroticized, with a palette of acid yellows, glowing pinks, and deep greens. Intricate details, oil painting texture.

About Wtewael: https://en.wikipedia.org/wiki/Joachim_Wtewael

Gemini won this week.
ChatGPT: https://drive.google.com/file/d/1qCQ7dRCNq1WPagkHt2HVFcBpdA7ihxvi/view
Gemini: https://drive.google.com/file/d/1-D00Hs5Z2FCWyM6BUtUrgZiFVQ90jQWb/view
Grok: https://drive.google.com/file/d/1CgPZPYQBLNxu54jsjjTGVfQpgM30DZEV/view
Perplexity: https://drive.google.com/file/d/1n_cKfzmVMV-JR45iFIaQ7-bobAJcdfB3/view

---
One may recall that this newsletter began with the pandemic.

The Pandemic Mitigation Collaborative (PMC) uses wastewater viral RNA levels to forecast COVID-19 rates over the next 4 weeks.
COVID rates are on the rise and typically continue to rise through the holiday season.
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/

Comments

Post a Comment