grand theft eigenvalue 🔆

skeets were never meant to live long

he's a ten but writes traditional and speaks with erhua

almost like there's an algorithm that makes the price go up when more people want it

the real problem is private equity buying all the oil to drive up the price

where are the nimbys explaining how oil prices aren't actually about supply constraints

the best lack all conviction. the second best got convicted

whichever side is least convenient for me right then tbh (left)

there's some of that. defense buyers have been interested in products that can gin up non-defense demand. in space, export controls generally have limited foreign competition as well as foreign demand. net effect for investors is not always favorable.

I'm ignorant here: how does accountability work with PMs? I feel strongly it's unacceptable for me to blame the LLM for bad code; are PMs the ones who take blame if the team ships bad code?

also,

3. the "inelastic labor supply" for the space workforce also kinda translates to a supply of idiosyncratically motivated investors. also hype and narrative. but i don't think these really explain new investor entrants; defense money is really the key these days imo.

i think it's really two things:

1. the 2010s let a bunch of people get in with exits built around _very_ speculative businesses. now the orgs exist and are looking for markets.

2. now there's defense as a customer of last/first resort. perceived as quality revenues.

this is a good thread. forking here to think about what makes space different, because it feels similar but is playing out differently.

i insulted entropy :)

both good options!

ha! tbh i'm getting tired of vibe prefixes... and it doesn't feel appropriate when there's tooling and systematic workflows. "agentic engineering" is a better fit than "vibe coding" imo but "i agentically engineered" is a mouthful

ok this is good this is progress. conjured for when it went well or produced something suitably beautiful/sinister, begot for the whelps i don't really want to look at but perform useful functions

So you're telling me that everyone now has access to a magical system that can correctly synthesize and answer questions about any doc or code that has ever been written by any person at your company, and deeply critique your design docs and code, but somehow junior engineers are *less* valuable

oh that's a good one. i definitely have projects where that's been the right term

i like created! i don't think i can consistently say i designed it -- i reviewed designs, i wrote specs and drafts, but the final design is often a composite of things i wrote and details the LLM included and i thought made sense

supervised is the other one i was leaning towards but...yeah, sticky

"how do PMs talk about code their teams produced" is maybe the closest fit. but there isn't an expectation there that the PM is accountable for the code in the same way the senior engineer/team presumably is. idk.

what's the right word for talking about things you produced with LLM support? like code. i orchestrate the production. i review it. i am accountable for it and its effects. but i wouldn't say i "wrote" it... would i? i guess typing is already a step removed from writing so maybe "wrote" still works?

claudatory was right there

who up signing their referee reports

computer and math also looks quite promising

never been a better time to get into business and finance :)

an editor once said, "I want to publish beautiful papers". if every one-shotted paper is beautiful then none of them will be

Journopoclypse! Yeah, na. I don't think so

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

valid haha

I don't recall exactly but ballpark maybe 67%?

love that this gives me the option to download my posterior