@metacurity@infosec.exchange2026-05-12 20:07:19
China Sought Access to Anthropic’s Newest A.I. The Answer Was No.
https://www.nytimes.com/2026/05/12/us/politics/china-ai-anthropic-openai-mythos-chatgpt.html
@grahamperrin@bsd.cafe2026-07-14 06:55:03
Two recent comments about Reddit's AI search:
<https://www.reddit.com/r/OutOfTheLoop/comments/1k76q1l/comment/ovp5xf4/>
― previously known as 'Reddit Answers'
<
@bibbleco@infosec.exchange2026-07-13 14:06:28
Just did very approximate calculations; with 10g of lithium per drone*, and assuming the Russian and Ukrainian drone efforts consume 5m units a year**, that'd be ~50 tons of lithium deposited into the environment per year, mostly within, let's say, 50km of the grey zone***. It's only about $20,000 / ton****, which means $1m of metal*****.
So to answer my own silly question: no, it won't be economically viable to scalp the top few inches of topsoil and extract the lithiu…
@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.
@thijs_lucas@norden.social2026-05-12 06:17:54
Die feine @… mit ihren feinen Stoffen würde niemanls die eigenen Leute bedrohen oder erschießen, um ihre Politik gegen geltendes Recht durchzusetzen. Sie rüstet dafür ausländische Milizen aus und lässt sie in internationalen Gewässern auf deutsche Schiffe schießen.
Diese Schiffe voller Menschenleben sind leer von Öl und Gas und kaum sc…
@tiotasram@kolektiva.social2026-07-12 16:28:07
Since it was relevant to a discussion I just had on here and is something most people probably haven't thought about much (unless you've taken one of a handful of philosophy classes), I thought I'd try to lay out a key piece of Descartes' Meditations (#philosophy
@Techmeme@techhub.social2026-05-12 10:35:47
Sources: Anthropic officials refused a Chinese think tank's request to change its stance and allow Beijing to access Mythos at a meeting in Singapore last month (New York Times)
https://www.nytimes.com…
@memeorandum@universeodon.com2026-05-11 21:05:38
Trump nominates Cameron Hamilton to lead FEMA, a year after he was fired from the role (Gabriela Aoun Angueira/Associated Press)
https://apnews.com/article/fema-cameron-hamilton-trump-disasters-navy-seals-e1ef0f6c81f6ea992a2213714f6743b1
http://www.memeorandum.com/260511/p99#a260511p99
@matzekult@chaos.social2026-05-12 12:02:03
Even though a lot of planets and species only appear once in #StarTrek, some leave a lasting impression; even if just for the memes. As is arguably the case with the subject of today’s #TrekTriviaTuesday question.
As always no googling and no spoiling the answer for others. Please boost a…
@aardrian@toot.cafe2026-07-10 15:42:46
