Xulrah

Every prediction that guy has made the last 5 years has turned out true. But that doesn’t mean they’ll continue being right. So instead I should trust the guy who’s been wrong every year because he has a PhD and his ideas sound normal.

No longer good reasons to start a startup: - have a good idea - want to change the world - want agency Good reasons to start a startup: - extraverted - previous jobs are already as prestigious as it gets - have a large following and want to monetize it - rejected from labs

Twitter people are hopeless at reasoning about rsi

Opus has been down for the last couple of hours and I am crippled

Cycle time from high status person makes phrase popular -> Twitter mocks low status person using that phrase to gain status is growing ever shorter

Why do you have to trick your laptop into thinking you want to print something to save it as a PDF

The cycle time before a good thing something becomes performative is rapidly approaching zero

Kind of insane how well free markets work

girls in sf date like value investors

there's a whole world of Asian girls in sf that become interested in you the day you start working at an ai lab

The best founders used to be indifferent to short-term prestige. Funny to think about that now

Genuine q, is there any deliberate coordination involved when everyone in tpot starts the same bit? @sporadica @eunifiedworld @creatine_cycle @personofswag

Great explanation of latent space

The intersection between young engineers that use vim and engineers with good product sense is null

Alex Wang has a deep seated resentment of most of most of humanity. This is his most public expression of that yet

A really good way to eval your API design is good is trying to have an LLM one-shot working code with only your function signatures in context.