it is indeed very good that there are open models. their training is still often opaque; what might they be optimized for? there are enough economists and behavioral scientists in tech that I could imagine an open model that somehow maximizes a profit function #linklog

RE: https://bsky.app/profile/did:plc:yw3qqfa7hrhmwbc6wyrq5u7f/post/3mhgplshf7k2t
a core thing this gets right: scarcity of positional goods is a conserved quantity, and time to compete for them is always finite. I think everything else -- the role of taste, understandable summaries -- follows #linklog

Journopoclypse!
Journopoclypse!

Yeah, na. I don't think so

Joshua Gans' Newsletter
"It’s also undeniable that SpaceX needs more than just Starlink to justify a $1.5T valuation, given that even its expected lead investment bank, Morgan Stanley, only thinks Starlink revenues will get to $126B in 2040." -- this is the sharpest thing i've read about the SpaceX IPO yet #linklog

TMF Associates blog » SpaceX’s...
TMF Associates blog » SpaceX’s Rorschach test

as much as I like to say that AI is a social science, I find this a reasonable argument that CS isn't quite there yet (and should get closer). on the other hand, mechanism designers' claims notwithstanding, economics could use more build-and-test if it's to build a real engineering culture #linklog

The Empiricism Gap in Computer...
The Empiricism Gap in Computer Science

So far, I have argued that there is a dissonance between, on the one hand, CS’s founding myths, curricula, and self-image, and, on the other hand, the modern production of knowledge in computer science.

Doomscrolling Babel
"We learn what generosity, vanity, or integrity look like by watching them play out in lived situations" -- This lens is more interesting to me lately than "Jane Austen, game theorist" (albeit complementary). Who is writing this kind of applied virtue ethics fiction particularly well today? #linklog

Jane Austen as Applied Moral P...
Jane Austen as Applied Moral Philosopher

Austen's fiction is not "about" romance. It's about character. And we need it now.

Knowledge Problem
I don't know that anyone has ever made a very good UX for sampling from high dimensional joint distributions, but etiquettes can work well in their native equilibria #linklog

"AI" is bad UX
"AI" is bad UX

teapot from the cover of Don Norman’s “The Design of Everyday Things” clumsily ‘shopped by me "AI" means bad UX There is an emergent strain of thought in...

Apperceptive by Sam
"The high cost of housing is the cost required to get the 15 millionth household to wait." in a sense many households are tied for 15 millionth #linklog

Are we underestimating who is ...
Are we underestimating who is poor?

Michael Green has caused a bit of a dust-up with a piece that was picked up by the Free Press, and which he followed up on at his substack newsletter.

Erdmann Housing Tracker
Hear me out: what if, instead of a #linklog that just keeps growing and I hardly ever look at again, there's just a little section on my homepage: "websites holding my attention these days".
I like this term "wash hire". data drift, particularly for firm-specific knowledge, seems a reasonable driver of wash hiring. #linklog

AI Replacement and Wash Hires
AI Replacement and Wash Hires

What could be causing cycles of hiring and firing in the labour market?

My weekly collection of cool stuff I've read on the web, Scraps, comes early today. We've got tic-tac-toe made with only CSS (as in, no html 🤯), some thoughts on LLMs, a way we can be better in the indieweb, an amazing in-progress technical reference manual and a whole bunch of other #indieweb #infosec #tech and more stuff to peruse. Some great stuff this week, #humanweb! Thanks for being awesome!

https://fyr.io/scrap/2025-07-18

#linklog #wrapup #weeklynotes

Scraps

Scrappy notes on cool and interesting stuff I've seen online from the last little while