Favorite Rejects 2023-03-18
At the state archive, checking out some bills. Here's the file given to the Governor when signing SB443 in 1968, the bill that authorizes BART to use land value tax to fund itself (the board declined and used sales tax instead to make up the inflation-caused budget issue 😑)
Wow, Russia is struggling to encircle a heavily fortified enemy city?
That sounds like an easy job for a few thousand elite paratroopers, I sure hope Russia didn't lose them all in a pointless operation during the early days of the wor
Had a meeting today at work where I pulled a coworker into a conference room to clear up a misunderstanding with another on a video call.
It was so much easier than trying to resolve the discussion over chat. I can see why data shows that people are more productive in the office
OK this is the weirdest “X is closer to Y than Z” I have seen
Chris Plante Show @ChrisPlanteShow
Philosophers have long debated whether “the hard problem of smart television” is solvable, even in principle, by any sort of physical system while others dismiss this as myth-making and superstition.
On a more abstract level, DSA’s embrace of the “we should negotiate every housing project as an individual ‘deal,’ and that deal is never good enough” framework has done more to dissuade me from flirting with socialism in the past few years than anything else
DSA Housing Justice Commission 🌹🌆 @dsa_housing
“Missing middle” can’t solve our problems because unlike Paris we don’t have a Haussmann to tear up neighborhoods of small buildings and replacement wholesale with “gentle density.”
We need to build high rises in the present while we wait for gentle density
Nia is looking for a full-time VC job✨⚡️🌙 @_johnsonator
It’s odd to me how many (seemingly?) practicing architects believe that bedroom window requirements are about fires. US codes do not contemplate window rescue the way foreign ones do – casement windows barely open, and they often look out onto inaccessible rear yards
The Hustle Architect @VitruviusGrind
A simple heuristic that could've predicted AI progress over the last 10 years:
- We can reach superhuman performance in bounded domains with explicit scoring systems (Go, Chess)
- We can reach ~human performance in areas with tons of real-world data (writing, reading, vision)
