@zydecopaws@pnw.zone2026-04-18 03:24:28
We’d finally get those flying cars we were promised.
#IfCartoonsRuledTheWorld
#HashTagGames
@chris@mstdn.chrisalemany.ca2026-04-19 16:28:49
Horrendous.Russia’s war machine on the brink of collapse. Only Trump’s sanctions relief is keeping the money and drones flowing at this point. Women and children stolen from Africa being used to manufacture.
“The Alabuga Special Economic Zone in Russia’s Tatarstan region is the primary production site for the Geran family. The factory operates around the clock and has sought thousands of workers, primarily young women and girls — some as young as 15 — recruited from Africa. About 200 African women, mostly between 18 and 22 years old, are employed at the facility, with documented reports of workers describing the conditions as a trap, with costs for accommodation, airfare, and Russian-language classes deducted from their wages. In July 2025, multiple reports — including a documentary by the Russian defense ministry’s own Zvezda channel — indicated that Russia was using children and teenagers to assemble the Shahed drones used to attack Ukraine.”
#russiaUkraineWar #russia #usa #TheAmericanFascist
@aredridel@kolektiva.social2026-04-14 14:22:42
So to follow up on this, I've caught it in action. Models, when quantized a bit, just do a bit more poorly with short contexts. Even going from f32 (as trained) to bf16 (as usually run) to q8 tends to do okay for "normal" context windows. And q4 you start feeling like "this model is a little stupid and gets stuck sometimes” (it is! It's just that it's still mostly careening about in the space of "plausible" most of the time. Not good guesswork, but still in the zone). With long contexts, the probability of parameters collapsing to zero are higher, so the more context the more likelihood you are to see brokenness.
And then at Q2 (2 bits per parameter) or Q1, the model falls apart completely. Parameters collapse to zero easily. You start seeing "all work and no play makes jack a dull boy” sorts of behavior, with intense and unscrutinized repetition, followed by a hard stop when it just stops working.
And quantization is a parameter that a model vendor can turn relatively easily. (they have to regenerate the model from the base with more quantization, but it's a data transformation on the order of running a terabyte through a straightforward and fast process, not like training).
If you have 1000 customers and enough equipment to handle the requests of 700, going from bf16 to q8 is a no-brainer. Suddenly you can handle the load and have a little spare capacity. They get worse results, probably pay the same per token (or they're on a subscription that hides the cost anyway so you are even freer to make trade-offs. There's a reason that subscription products are kinda poorly described.)
It's also possible for them to vary this across a day: use models during quieter periods? Maybe you get an instance running a bf16 quantization. If you use it during a high use period? You get a Q4 model.
Or intelligent routing is possible. No idea if anyone is doing this, but if they monitor what you send a bit, and you generally shoot for an expensive model for simple requests? They could totally substitute a highly quantized version of the model to answer the question.
There are •so many tricks• that can be pulled here. Some of them very reasonable to make, some of them treading into outright misleading or fraudulent, and it's weirdly hard to draw the line between them.
@_tillwe_@mastodon.social2026-02-12 08:09:25
@der_raddler@dresden.network2026-04-13 16:09:45
Gerade noch schnell einen Beitrag zum heutigen #FotoVorschlag 'ziemlich schräg' hochgeladen.
https://photo.dresden.network/p/der_raddler/949334084161225414…
@Mediagazer@mstdn.social2026-04-02 21:05:53
A photojournalist and the Reporters Committee for Freedom of the Press sue the FAA over a ban on flying drones within 3,000 feet of DHS buildings and vehicles (Matthew Gault/404 Media)
https://www.404media.co/journalist-sues-faa-over-drone-…
@memeorandum@universeodon.com2026-04-08 10:40:44
IDF Strikes Kill 8 in Lebanon After Netanyahu Says Deal Excludes Hezbollah (Zen Reading/Reuters)
https://www.haaretz.com/israel-news/israel-security/2026-04-08/ty-article-live/mda-three-children-lightly-injured-in-negev-following-iranian-rocket-fire/0000019d-6ad1-d940-a5bd-fedf1e030000
http://www.memeorandum.com/260408/p7#a260408p7
@Simone21@mastodon.social2026-04-06 12:09:41
@zydecopaws@pnw.zone2026-04-07 01:20:24
A magical device that allows all knowledge to be available to even the least of man shall be used to watch images of felines.
#BoringNostradamusProphecies
#HashTagGames