hey wasn't this the same company that made a beautiful shiny "research" post about how AI evals should include error bars or something like that. or did they decide the CLT didn't apply here

individual reporting for post-deployment evals — a little manifesto (& new preprints!)

tldr: end users have unique insights about how deployed systems are failing; we should figure out how to translate their experiences into formal evaluations of those systems.

help ..

where are the bullshit "x% of experts believe" polls when you need them lol

im ngl i think this kinda just means u are stupid

... didn't we just talk about this ...

were you born yesterday

i just know these people would have been the biggest fans of japanese internment

sorry, is this supposed to be interesting or novel or surprising???

this is idiotic for so many reasons

i am NOT seeing enough trolling about this "did you know that sometimes you can use the CLT for evals" be so for real right now

poetry and manifestos in an anarchist hackerspace.. still so psyched about the turnout for kernel magazine x lit crawl :^D

umm... nice i guess 👍

cant miss april .. day after perfect frickin day

it’s march

conference deadlines? we don’t know her (happy 元宵节)

this is maybe kinda nice. think i’m just gonna spam about how much i love grad school here