Navier–Stokes: “… the [agent] group that produced the Navier–Stokes resolution involved on the order of 10,000 concurrent agents”

https://openai.com/index/navier-stokes-solution/

The explanation of what they found is readable. Model beyond Astra. 88 hours 300 billion tokens.

“When a further trained version of our internal model became available over the course of the effort, we updated our agents to that model.”

“Researchers found 258 different genetic variants linked to extroversion … the trait was also linked to neurodevelopmental conditions such as A.D.H.D.

https://archive.ph/20260903212043/https://www.nytimes.com/2026/09/03/science/genes-personality-traits.html#selection-4273.0-4273.406

We probably have to come up with new names for personality traits that line up better with the genes that influence temperament. Also, sociobiology returns…

(Article and researchers speak of personality, but I think they are actually studying temperament – of which personality is an adult observable consequence.)

Tsai: “Normally, I update around January. With Tahoe I waited until July … considering updating to Golden Gate soon after its release. “

https://mjtsai.com/blog/2026/09/03/preparing-to-update-to-golden-gate/

I usually update around March (Michael is a dev so he had to update). This year I will update around November. It’s expensive because I’ll need to give up on my old Synology TM backup. (I’ll probably drop TM backup altogether for a while.)

“DSM is an official classification of mental disorders, but it neither reflects new scientific developments nor established clinical wisdom”

https://www.psychiatrymargins.com/p/postpartum-psychosis-and-diagnostic

DSM is obscure, extremely important, and widely despised by thoughtful psychiatrists.

I believe it is also a critical funding source for the professional society.

Very insider post that is worth a look.

Ethan Mollick on instability arising from exponential growth in ai abilities

https://www.oneusefulthing.org/p/the-twilight-of-the-chatbots

“instability is what happens when institutions that move at the speed of people … try to track a capability curve that is very much not human in nature. And as long as we are on some sort of exponential, and for as long as it lasts, the gap only widens.”

Too much change too fast. And pre-ai automation/globalization change was already too fast for modern societies to adjust to.

Caveat: Leading models still can’t do a reasonably tax return. So there some big holes and they are not closing exponentially …

Culture shift and the decline of Lizzo

https://www.theatlantic.com/culture/2026/07/music-flop-era/687785/?lctg=6050e8354c8a1e4095fc2896

There’s a lot of chance to fame and glory, and few stay near the top. The interesting part is tying pop star decline to changing culture …

“This era of the internet is becoming stale, and the trajectory it set culture upon is reaching a terminal point. After years of explosive expansion, U.S. growth for Spotify and TikTok has shown signs of slowing”

I don’t know where the culture goes. Might was well try to predict fashion. But it does feel in motion again.

“data-driven investigation into Democratic fundraising, documenting how a network of consultants, vendors, and party committees extracts money from elderly donors through manipulative digital fundraising.”

https://data4democracy.substack.com/p/with-fear-and-favor?r=7n61q

“the DCCC and DSCC responded with a cease-and-desist letter from the Elias Law Group demanding that the Times not publish.”

NYT caved

“As predicted, rent stabilization policies and politics in both St. Paul and Minneapolis have dramatically reduced new housing construction”

https://www.minnpost.com/cityscape/2026/06/lowlights-and-glimmers-of-optimism-from-a-new-st-paul-housing-study/

St Paul rent control approach is kind of suicidal.

“St. Paul saw fewer apartments built last year than Lansing, Michigan, a city that’s been shrinking for a half-century.”