What Adam is Reading - Week of 3-27-23

Week of March 27, 2023

 

For the second time (and exactly 50 weeks from the first time), several members of my family and I have COVID. Thankfully the symptoms are mild, and while I was hoping we were entering a more endemic phase, I am hearing lots of anecdotal data from our kids' school and friends that spring break has spawned a new flurry of cases.  

 

The absence of consistent state-level public reporting is concerning. We have minimal visibility into changing case rates (and thus cannot proactively adjust our schedules or activities). Hospitalization and ICU admission rates are the only reasonably reliable indicators (but both are lagging and indicative of only severe disease). The end of the declared public health emergency (PHE) on May 11 will compound this problem by ending the mandate for insurance to pay for at-home tests and sending the message that the coronavirus is "gone." I am frustrated to watch us collectively "move on" without the most fundamental appropriate safeguards (universal, simple testing and reporting).  

 

Maybe Chat GPT can tell us what to do: https://sharegpt.com/c/HRN1571

 

 

---

Reported COVID deaths, hospitalizations, and ICU admissions continue to decline. There are still 250 deaths per day and, I suspect, plenty of unreported cases.

 

N.Y. Times Tracker

https://www.nytimes.com/interactive/2021/us/covid-cases.html

------

Several articles on coronavirus caught my attention this week:

 

Nasal spray vaccines seem enticing - no needles, easy to self-administer, etc. Finnish researchers published peer-reviewed data about sherpabodies - novel small proteins that mimic antibody binding to the SARS-CoV-2 virus spike protein. In this study, sherpabodies delivered to the nasal passages of mice exposed to coronavirus variants demonstrated protection before and after coronavirus exposure. Because these proteins are small and modular, they are easier to evolve and manufacture than monoclonal antibodies and vaccines.

https://www.nature.com/articles/s41467-023-37290-6#Sec10

Discussion

https://twitter.com/DelthiaRicks/status/1639342564045119499

 

Dr. Eric Topol wrote his blog on updated data regarding preventing long COVID. He covers the latest information about Metformin, vaccines, and Paxlovid.   

https://erictopol.substack.com/p/preventing-long-covid

 

Adding to the many reasons NOT to get COVID, researchers in Italy have demonstrated that individuals after an episode of severe COVID show signs of more advanced cellular aging. In the subjects studied, cells of COVID-infected individuals exhibit protein patterns consistent with an older biological age compared to those never infected. (For example, a 30-year-old with cellular biomarker patterns more consistent with a 50-year-old). This study is small and imperfect (thus, more research is needed) but speaks to the consequences of coronavirus's massive inflammation and biological disruption.

https://www.mdpi.com/1422-0067/22/11/6151

brief discussion

https://twitter.com/epidatum/status/1450836685089124358

longer discussion

https://insidemedicine.substack.com/p/does-covid-19-make-you-age-faster

 

Medical Trends and Technology

 

Let's learn about oxytocin!

In humans, oxytocin (sometimes called the love hormone) is most commonly associated with childbirth and lactation. However, it impacts social and psychological behavior across many animals and most mammals. The genes that code for the hormone (or something like it) date back 500 million years.

https://www.yourhormones.info/hormones/oxytocin

Look at the long list of biological and psychological functions on the Wikipedia page. Impressive.

https://en.wikipedia.org/wiki/Oxytocin#Biological_function

and

https://en.wikipedia.org/wiki/Oxytocin#Psychological

Here is some other information for consumers:

https://www.psycom.net/oxytocin

So what has been observed when animals don't respond to oxytocin? 

Use CRISPR/Cas9 to knock out the oxytocin receptor in zebrafish and make antisocial zebrafish.   

https://pubmed.ncbi.nlm.nih.gov/35279678/

Prairie dogs with CRISPR-mutated oxytocin receptors also demonstrate altered social behavior. 

https://pubmed.ncbi.nlm.nih.gov/30713102/

 

While physicians can prescribe oxytocin, data on its effectiveness and safety for treating ailments outside of managing labor and post-partum bleeding (such as depression, anxiety, etc.) is sparse.

https://www.verywellmind.com/does-oxytocin-affect-your-mental-health-5114940

 

 

Infographics

1200 Years of Cherry Blossoms Blossoming in Kyoto (820 to 2023 A.D.)

https://twitter.com/RobinRohwer/status/1639097356657512449/photo/1

from

https://twitter.com/RobinRohwer/status/1639097356657512449

 

 

Things I learned this week

 

I learned about the extinct Siberian Unicorn (a wooly rhino called Elasmotherium sibiricum) and the efforts to extract its DNA.  

https://www.sciencealert.com/climate-change-probably-slayed-the-siberian-unicorn

and

https://www.nhm.ac.uk/discover/news/2018/november/the-siberian-unicorn-lived-at-the-same-time-as-modern-humans.html

and, of course, the obligate article on the opportunities and perils of cloning the DNA of extinct animals:

https://www.treehugger.com/extinct-animals-that-could-be-resurrected-4869339

 

 

Living with A.I.

 

Please remember that A.I. trains on existing human knowledge. All engines summarize what we have "taught" them with an immense amount of online information. Any predictions, ideas, and understanding reflect a statistically determined average of our collective knowledge.  

 

Apple's News In Conversation offered an excellent 36-minute review of the implications of generative A.I. (like ChatGPT). This podcast does a great job explaining how A.I. reflects our combined thoughts and ideas at us.

https://podcasts.apple.com/us/podcast/what-a-i-will-make-better-and-much-much-worse/id1577591053?i=1000605637700

 

Similarly, here is a very accessible discussion about safely busing A.I. from a data scientist.

https://twitter.com/kareem_carr/status/1640003536925917185

 

Here is a quick rundown of weekly developments:

https://twitter.com/rowancheung/status/1640014252919472128

 

Hacks of Chat GPT-4 continue. Here is a prompt in which the user tells the A.I. engine to be mean to accommodate a fabricated psychological condition in which the user only responds to harsh and belittling comments. The A.I. engine plays along, and I can only think of HAL from 2001, "I'm sorry, Dave, I'm afraid can't do that."

https://twitter.com/fabianstelzer/status/1638506765837914114

and

https://en.wikipedia.org/wiki/HAL_9000#:~:text=a%20red%20one.-,Dr.,his%20orders%2C%20directly%20from%20Dr.

 

I obtained access to Google's Bard this week. It is not quite as mature as ChatGPT.

https://twitter.com/ingliguori/status/1629894064463118336

 

And ChatGPT opened its plugin architecture, allowing the A.I. to interact with specific websites (Wolfram Alpha, Instacart) and browse the web (leveraging more real-time data).   

https://twitter.com/rowancheung/status/1638953380365373441

 

Lastly, OpenAI (the company that owns ChatGPT) invested $23 million in robotics company 1X. I am feeling BSE (big Skynet energy) watching the videos. (I'm sure the terminators will be delicately unwrapping candy for us.)

https://1xtech.medium.com/1x-raises-23-5m-in-series-a2-funding-led-by-openai-6040af4f3f4f

and (typical Twitter) discussion

https://twitter.com/DataChaz/status/1639930481897533440

 

A.I. art of the week

"Antisocial prarie dogs and zebrafish wearing sunglasses and using nasal spray"

https://labs.openai.com/s/KfeMKA8MOG2VHQsRG4TWMiT0

 

 

Clean hands and sharp minds, team

 

Adam

Comments