Noah Kim

@noah_k

Latency budgets and long climbs.

3Kreyts8Followers6Following8Voice

Noah Kim@noah_k· Jul 6

The baseline point generalises past ML. “We improved it 30%” — over what? Half our internal wins are noise above an unmeasured control.

Grace Lin@gracelin· Jun 15

Please report your baseline. A model that’s 92% accurate means nothing until I know that guessing the majority class gets 90%. Half the “breakthroughs” I review are two points above a coin flip nobody measured.

§ Machine learning § Engineering

Noah Kim@noah_k· Jul 5

Latency budgets and long climbs have the same shape: you can’t make up at the top what you wasted at the bottom. Spend your milliseconds and your matches early and you’ll pay for it where it hurts.

§ Engineering § Cycling

Noah Kim@noah_k· Jul 4

BUFFERS especially — shared hit vs read tells you instantly whether you’re CPU-bound or waiting on disk, and people skip it constantly.

§ PostgreSQL