Mostly true only if "video games" count as a higher dimension.
codexarcanum
joined 3 months ago
I love the idea of "From Drag til Dawn" and will have to check this one out, sounds right up our alley!
His exact tweet came up in another post of this story. He said we should eliminate Palestine and Isreal, which is that one step too far. Today's Nazi's actually can't be openly antisemetic because the genocidal apartheid state of Isreal is their closest international ally.
Deny. Defend. Depose. Disavow.
Goodbye!
(What an amusing outcome!)
This rules! I used to scrobble all my shit, time to get back to it!
Someone suddenly looks thoughtful while they chew, "To be delicious, is to be defiant" they mutter, entranced. Others begin to join in, a chant is growing, "Delicious! Defiant!"
who invited you here? who's party is this? what do they believe?!
Showing up at the function to find a life-sized recreation of George Floyd being murdered, rendered in cake. Everyone says that chocolate Floyd is delicious but the white almond officer cake is somehow both dry and a little undercooked.
I'm surprised the base layer is white fondant and not like a full color image of Clinton and Harris hugging in their best pant suits right before losing another layup election.
Agreed with @Breve@pawb.social's comment: not a good article.
First off, this article is just a link and a short summary of a much longer blog post of the Qualys AI security company. So this isn't a standard industry or academic benchmark, this is something Qualys is using as part of their sales strategy for AI consulting services.
Which, frankly, I would avoid if this article is indicative of how they work.
They tested 1 model. One. Insane. DeepSeek released about a half-dozen distillation of llama and qwep2: 3B, 7B, 8B, 14B, 32B, and the full 671B model. As an AI-expert security company, why would you test one of the weakest distillations only? Was it really so expensive to rent some time on the cloud to test out the other models? Not even the 32B which should be well within most hobbyist ability and budget to run on cloud and test out.
The full article is such obvious SEO bait, I'd be surprised if AI didn't help write it. Just goes on and on and on talking about why AI security is important, why these tests are important, why its important each single test that it failed. But ultimately, no matter how many words they write, the point that they only test 1 very weak version of the model makes the whole thing pointless except as a way to tie the name Qualys to DeepSeek and thus lure in more gullible rubes. Like did they only test 8B because that's cheap enough for them to run as a service for their business?
Anyway, I've previously written here on Lemmy about how Deepseek distilations are so easy to break than I almost think they're buggy and not just insufficient in this regard. I'd really want to see this kind of analysis done on the full model, and obviously also on the other big LLMs they could test to compare against. I wouldn't even trust the 8B models for unsupervised work. It's certainly not reliable and non-hallucenatory enough for things like content filtering or expert system usage, no 8B-level model I've tested is. So very concerning that it seems like that's the level of service they're pushing.
"Yeah, yeah, ung, that's what daddy likes, mmmmmmmmmm"
"Oh great, I'm glad you're enjoying the album!"
The new Prince of Persia is the worst for this! There's no auto-cloud save, so you have to manually manage uploading and downloading to the one cloud save slot between your three on-device slots.
And no matter what, no matter how long it's been, it asks about it, like: "Your last save was 0 minutes ago, as you sure you want to exit?"