Viewing profile — hoosieree

"send me your prompts instead"

The purpose of a system is what it does.

People don't respect the chasm between "obviously no mistakes" and "no obvious mistakes".

"You're right, that was your battleship. My bad! Would you like me to launch rockets at another target?"

Also water. You will be rationed, OpenAI will not. https://www.newstarget.com/2025-08-02-texas-ai-data-centers-...

This should go without saying but unfortunately it really doesn't these days: This kind of corporate behavior is bad and will end up hurting somebody . If we're lucky the fallout w…

0d (parked) - null program easy 1d (train) - easy. just one lever 2d (car) - hard. super hard. why is it so crowded? who thought this was a good idea? you let teenagers do this? 2.…

I found building musl easier than uClibc, Xephyr, newlib, or (especially) glibc. Glibc is an Ouroboros of hacks at build time.

Array languages don't just have shorter tokens, they have fewer of them . A small token set is practical because your language's users can only keep a finite number of things in th…

I ended up with Emacs and org-mode. One of my friends used Vim and Pandoc to the same effect. I didn't quite have to write my own tool but each new LaTeX template foisted on me was…

Or a 4-hour workweek.

And currently trying in the US too: https://apnews.com/article/ai-regulation-state-moratorium-co... They want to ban states from imposing their own regulations on AI.

AI also happens to be a perfect scapegoat: CEOs who over-hired get to shift the blame to this faceless boogeyman, and (bonus!) new hires are more desperate/willing to accept worse …

I never before considered that a password generator could be fun. My mind has been changed. Now I believe they should be fun.

The error in this argument is that electricity is real .

They missed one: def agent(data, predicate, fns): ps = map(predicate, data) fs = map(lambda x:fns[x], ps) return map(fs, data) Basically you want to apply one of N functions to eac…

I like overloading most of the time but prefer specialized syntax for the various conditional idioms. Encapsulating it all in one keyword makes "if" harder to parse.

The main point I was trying to make is that adding the prompt "think carefully" moves the model toward the "riddle" vector space, which means it will draw tokens from there instead…

My classifier is not very accurate: is_trick(question) # 50% accurate To make the client happy, I improved it: is_trick(question, label) # 100% accurate But the client still isn't …

LLMs can't hallucinate. They generate the next most likely token in a sequence. Whether that sequence matches any kind of objective truth is orthogonal to how models work. I suppos…

Video autotune.

Apparently they can haul 50+ sheets of 4x8 plywood with the door closed (and without removing the seats).

In the sci-fi series The Expanse, there's a Bezos-style rich tycoon character who flaunts his wealth by purposefully not getting hair treatments and instead allows his male pattern…

Heard of the Jevons paradox?

I used to work with psychology researchers conducting experiments with wearable cameras. Anything involving human subjects needed IRB approval, informed consent, ethics review, etc…