🧵 1/ AI may break the internet because it reduces the effort to produce good content, and therefore EFFORT no longer acts as a reliable signal of quality. There's an interesting economics back story if anyone is interested below.

wired.com/story/ai-slop-… 2/ First, the Internet is an information "Market for Lemons."

When used car lots were new in the 50s-60s, there was no way to tell whether you were buying a bad car (a lemon), or a good one (a peach). There was no Carfax or even standardized VINs.

sfu.ca/~wainwrig/Econ…

🧵I wanted to contribute to the incredible Corsi-Rosenthal Box community, and I figured out how to do 2 things:

1. Getting more granular speed control
2. Making it smart - having it adjust speed based on high CO2 or PM levels. Full post for sharing:

theair.substack.com/p/how-to-make-… 2/ CR Boxes just work so well - in recent research they can be 2-4x better than some HEPA filters at less than half the cost.

1/ An article and 🧵 about lead and why cleaning the air saves $$$.

New research involving NASCAR races puts the external cost of a *single gram* of airborne lead at $1,100.

Whole piece is linked, but highlights are here. Sharing is much appreciated!

theair.substack.com/p/nascar-leade… 2/ To understand the story of lead, it helps to understand just how useful it is. It melts at low heat and doesn't corrode, so you can work it with little technology. As such it was used by virtually every ancient society.

"People do what other people do."

People take cues from those around them as to what is safe / acceptable.

Covid is deadly, but deaths are still rare and time-delayed such that the direct consequences of behavior are abstracted, especially if you are in a conservative bubble.

https://twitter.com/celinegounder/status/1470769782047879175

I find it personally infuriating, because I have close family who aren't vaccinated and who genuinely believe that the mRNA vaccines are dangerous, but since the effects of high covid load are hidden behind hospital doors, it perversely allows people to "move on."

Endemic flu is (a lot) worse than you think, and that suggests that we are significantly underinvesting in *long-term* solutions like healthier buildings and rapid testing.

A🧵and article exploring what flu says about endemic Covid 1/

Read here: theair.substack.com/p/what-if-the-… 2/ The sense that Covid is just going to go away at some point is rampant, but it's also counterproductive, because it encourages short term thinking. "Endemic" Covid is going to take investment to manage, and this doesn't appear to have set in.

Adding a little lithium to municipal water supplies might reduce violent crime and suicide.

There are several studies that look at naturally occurring levels of lithium. This study from Japan shows detectably lower crime in areas with more Li.

cambridge.org/core/journals/… Another study from Texas showed a reduction in criminal behavior and hard drug use, but no effect on things like marijuana or DUI. pubmed.ncbi.nlm.nih.gov/1699579/

Why admitting Covid is Airborne is hard. A 🧵 and 6k word piece on the curious history of airborne transmission.

Evidence that Covid spreads through the air is strong but airborne protections are lacking. I wanted to understand why, and was surprised...
theair.substack.com/p/why-covid-is… It turns out that the default advice for most diseases is the same as we got for Covid - wash hands, keep distance, cough into your arm. That's because fomites and close contact are believed to be responsible for most transmission. But why is this so if airborne evidence exists?

CO2 Monitors work! A🧵

1. Sent one in daughter's backpack.
2. Identified 4k PPM CO2 in afternoons.
3. Had district investigate - they did -repaired 2 air handlers.
4. Rebreathed air % dropped 90%!

Thanks @Poppendieck @DavidElfstrom @WBahnfleth

Please share so others know! This is what things looked like before. CO2 gradually built up during the day. Mornings have gotten higher due to band and I still need to figure something out there, but this is a HUGE improvement in the afternoon.

https://twitter.com/jeremychrysler/status/1427812197657366535?s=20

CO2 at a local school peaked at 4,385ppm and was never below 3k in afternoon. I don’t want to overreact but this is more than twice as bad as I thought it would be. Acc to @DavidElfstrom at this level 9-10% of every breath was recently exhaled by someone else. Thoughts? One bit of good news is that the school bus ride from cross country doesn’t show up between 7:25 and 7:45. You can see the ride in the car to cross country an hour earlier, and the fast return to normal at the end of the day.

1/ Since the beginning of the Pandemic, prominent authorities have dismissed "aerosol" transmission. This stems from a long standing idea that droplets larger than 5µm settle quickly, within 1-2m, so personal contact (droplet) spread has been considered primary. 2/ But aerosol experts like @linseymarr @jljcolorado @kprather88 and @ShellyMBoulder have jumped in to point out that even 10µm droplets stay in the air for much longer, and Paul Baron's pres to CDC suggests the same.
drive.google.com/file/d/1T4Rb-H… and cdc.gov/niosh/topics/a…

Thread:

1/ Without good, consistent antibody testing, we don't know the total number of people who've had COVID-19 in a given area.

But we do know that New York has 90 times more deaths / person than Hawaii. 2/ By using deaths per person, we can estimate total infections by plugging in several IFR estimates and working backwards.

If you think the IFR is low, you can plug it in (as I have) and figure out what the infection rate has to be for that fatality rate to be true.