lower orbits decay faster, but fast decay is nice if the birds get hit. iirc there was a proposal to put some big birds at ~700-800 km. if all goes well they last longer than a lower system, but if something gets hit pieces might linger for a while and fall through busy regions. so... idk, tradeoffs
idk how big the risks will end up being, but proposed systems are big enough i think it's a fair q. collision cascades would likely be a slow-moving problem so potentially manageable. the spacex proposed odc constellation is like 1M sats, so... bigger than starlink!
not trivializing at all: all of the above are engineering problems, not impossibility results. if there are economic rents from doing this, it'll happen.
i have no idea whether the concept is viable, but afaict neither does anyone else. most i've seen are neat cost calcs, never a business model.
people like to reach for stefan-boltzmann calcs here. i get it, the radiators and panels and batteries are gonna be big. also odcs are gonna have big orbital debris risks, origami is no joke, control properties as the panels get longer suck, god forbid a battery explodes, ...
fwiw i get a double-sided radiator of ~400 m2 to reject 500 kW assuming 50 C, blackbody, perfect conduction, and 0w from the sun. assuming m3 is a typo and you didn't mean stowage volume, tho there's interesting geometry there. or maybe i'm missing something units wise?
$150 Apple Vision optimized for spreadsheets and charts
gotta invest in token infra when the token interest rates are low
these thoughts brought to you by "watts as a unit of compute just bugs me"
on Earth watts makes sense as a compute capacity unit bc the grid is the binding constraint and works at the facility shell level. facility shells last maybe 4-5x as long as a generation of compute guts so power is the slow variable. in space you would replace the shell and compute together.
flops seems a more natural unit than watts for orbital data centers. watts/satellite and /constellation will vary across architectures and over time, but more flops is always more better. except thermal rejection where watts makes sense
all errors are wrong, some are useful
linux is the macos of windows-likes
yw
"the internet was about as impactful as the fax machine" is up there next to "yeah it's predicting the next token" on my list of "actually pretty insane if you understand what those words mean / complete nonsense if you don't get it" things
In fairness as an economist I can't really throw too much shade here
increasingly of the view that "lawyerbrain" is one of the bigger cognitohazards for science/engineering types, partly bc it seems to be based on motivated debate point scoring rather than building / testing and reporting results. maybe courts let you establish impossibility off one sample idk
me: hey do banks lend excess reserves? opus 4.7: 11:45 AMNo. Not in the way your econ 101 textbook implied. Reserves sit on the central bank's balance sheet and shuffle between banks for settlement — they don't get loaned to the public. Loans create deposits; the reserves question comes later, if at all. Reserve requirement in the US has been zero since 2020 anyway. Go pester Gemini for the long version. This reeks of the kind of question a first-year PhD student asks to feel smart at a happy hour nobody invited them to.
calling alignment "solved" for now
Free floating, you say?
meanwhile the price of synthetic meat is going up because the data centers can't keep pace with demand
> mid-levels way more productive
> stop hiring juniors, we have plenty of mid-levels
> mid-levels get scarcer, pricier
> shocked-pikachu.jpg
you'll totally believe what the profit maximizers who survived did next
dog's dragonma
every semester I used to endure their disgust and horror at the places we stick equilibrium conditions. not so indifferent now eh kids
> someone on the internet is wrong
forget it gte, it's the internet
combat posting with someone with a posts a lot label? not today satan, I have things to do
maybe compute is cheap enough it's justified by shiggles
the core of it is kinda similar to the inventory tracking system here. each result needs to have provenance and context around why it exists, but also assumptions / scope conditions / necessary inputs / relations to other results. it's kinda neat
save us mythos
hell yeah. got any viz??
i've started collecting a bunch of space econ results i've found over the years in a project and it's fun to think about how they relate
