Keep on to blur preview images; turn off to show them clearly
like lowkey don’t underrate that if you’re the 7th engineer on apple notes that every feature you ship goes to a product a billion people use everyday. that’s lit as fuck
dei ex machina @openai, past: posttraining o3/4o, sora 1 & 2, applied research
my take nowadays is that lots of incredibly talented smart people waste years of their life floundering independently or being generally ineffective, mostly due to lack of experience/knowledge, when they should be inserting themselves into institutional power structures and quickly just dominating within them (given low talent bars). i’ve seen many friends join large established or developing companies and instantly climb within. most business, impact, and important work is power-law distributed. but then the flipside is you really shouldn’t get stuck too long either. jump into an org, make a large impact, get experience, then use your newfound knowledge of how the industry/world works to go build something new.
like lowkey don’t underrate that if you’re the 7th engineer on apple notes that every feature you ship goes to a product a billion people use everyday. that’s lit as fuck
i’m deeply grateful for the unbelievable physical and psychological security i have. i’ve never worried about my home, my town, or my country being invaded and pillaged. it is an incredible privilege to be excused from the at minimum mild ever-present anxiety of fearing for your wellbeing that most humans in history experienced every day.
dei ex machina @openai, past: posttraining o3/4o, sora 1 & 2, applied research
being a financier is so in right now. nat friedman, jeffrey epstein, tyler cowen. ever since i was young i wanted to grow up to be a financier
dei ex machina @openai, past: posttraining o3/4o, sora 1 & 2, applied research
hotel room notepads have some strange psychic energy which makes any idea inscribed on them at least twice as good
extended stay, holiday inn, marriott, hilton. the worse the hotel the better the notepad
the common response is that the models just want to learn and transformers are the thing we know how to scale really well. but this begs the question why transformers are the one thing we know how to scale? it’s not just sunk cost
dei ex machina @openai, past: posttraining o3/4o, sora 1 & 2, applied research