@w7voa one is metric, the other is imperial…is that normal?

me: “Zoom is launching new AI Avatars that can attend meetings for you, even making decisions on your behalf” 8yo: “can it see forward in time” me: “no” 8yo: <scowls> “lame” https://news.zoom.us/zoomtopia-2024-unveiling-ai-first-work-platform-innovations/

i have a social event to go to tonight, and i realize i need to show up late, bc that’s what people do, but it is PAINFUL to intentionally be late

i currently have a component that i wrote that i wish would die in fire, but i can’t figure out how to replace it

can't do **this** with a junior engineer

yooo tyro looks hot. i've never seen it before (and that's why we read open source code) A #pythonhttps://github.com/brentyi/tyro

When Hinton received the call that he won the Nobel prize in Physics, he thought it was a prank call, which is just about the most appropriate reaction possible for the guy who made deep learning

coding bootcamps were the canary in the coal mine for #AI coding if time-to-demo is good enough of a metric to rationalize large bootcamps, then your most of the way to a culture where poorly deployed AI coding tools are going to be accepted with open arms

when testing LLM apps, i find myself thinking like a software engineer, building test sets based on edge cases rather than normal distributions. anyone else do that?

it’s always annoying when you ask an internet full of engineers for an opinion and no one has one pro tip: instead, always share a bad opinion