andrew
banner
whltexbread.com
andrew
@whltexbread.com
I routinely fail Turing tests.
but can it kirkland?!?!?!?!??!??
January 12, 2026 at 11:06 PM
Let me know what you decide!
January 12, 2026 at 4:08 PM
I’m interested in how they measure this—Google attached Gemini to every search result, right?
January 8, 2026 at 4:27 PM
Agreeing with previous reply. Override this veto AND make a bill to regulate ICE or abolish it via appropriations. AND pass war powers act.
January 8, 2026 at 4:25 PM
piling on—don’t fall behind Schumer again when he inevitably chickens out
January 7, 2026 at 9:33 PM
fwiw, this isn’t particularly controversial, especially in academic circles. people who know how these things work know their limitations and generally agree with him.

(ask my friends, i’ve been saying this for years)
January 2, 2026 at 11:07 PM
you know, universal access to industrial-strength eliza was kind of a genius marketing move. the world didn’t/doesn’t have the necessary context to evaluate the tool on its merits.
January 2, 2026 at 4:02 PM
i agree with you, but i’m wondering if it’s one of those things where somebody defined a term in some paper and people cited it a bunch but never questioned that specific term

i am happy that there are some people who are not _just_ trying to scale transformers to infinity though
January 2, 2026 at 3:01 PM
☹️
December 30, 2025 at 11:16 PM
well, that’s something!
December 30, 2025 at 9:17 PM
tabor is so insidious. it is _impossible_ to have productive conversations about repealing it. why does nobody care that no other state has anything like it?
December 30, 2025 at 9:13 PM
really?!
December 10, 2025 at 10:08 PM
I like equinox!!!
December 9, 2025 at 8:23 PM
this is absolutely solvable. there’s a tricky trade off between the length of the power measurement window and the “smoothness” of the normalization.

but it’s nothing a two-pass algorithm and a speech activity predictor couldn’t solve.
December 4, 2025 at 3:32 PM