“The Supreme Court says we can use political partisanship to do congressional redistricting,” said state Rep. Todd Hunter, a Corpus Christi Republican and the bill’s author. “We will push forward.”

WTF.

We are so screwed…

Even if I had infinite money, I just don’t picture getting much satisfaction from that…

Makes sense, I guess, but still.

Unreliable user profiling is just what mods need… yep.

Why don’t they sick it on finding bots instead? Oh, right, because that would cost them user numbers and make them look bad…

What about auto flagging posts that break sub rules for real mods to check? Ah, well, they can’t monetize those outputs can they?

Do any of y’all pay for OnlyFans?

Like, I am single and horny/lonely, but… I don’t get it, even theoretically. What’s the appeal of viewing some performer through a one-way digital window over the deluge of free content?

And I get that some would be into that, but (apparently) a significant fraction of the planet? That I don’t understand.

And activity is rising as far as I can tell, look at the graphs at the bottom.

Many VPN companies post audits, and build up reputations. Not that I’d recommend it specificlly (since I only use it for a lifetime subscription I bought in a sale), but FastestVPN advertises the former.

…I guess it depends what you’re doing, too. If you’re, like, a government whistleblower, you might want to look into Mullad layered with something else instead of a more traditional commercial provider.

I wonder what’s up with that number of servers decline.

Consolidation? Which is fine to that extent.

I’d wager many experimented with hosting a younger Lemmy, and hosts who couldn’t sustain it got shaken out over time.

I wouldn’t assume the rise in posts/comments is all bots, either. I don’t have anyone blocked, and it doesn’t feel overrun to me.

The reduction in coverage was most pronounced before primary elections.

The reduction in staff covering politics made it harder for voters to differentiate between moderates and extremists in partisan primaries, and allowed extreme candidates to do better than they did before.

This makes sense, and it explains a lot, actually.

And to be clear, it’s not just craigslist as a culprit here, but it’s such a controlled A/B test that the effects are reliably measurable.

In the original paper, they also observed reduced turnout for House/Senate elections (which the article didn’t emphasize as much, but is defininitely there): https://academic.oup.com/restud/article/92/3/1738/7665573?login=false#517516514

The camera button confuses the heck out of my Mom (a new iPhone owner).

I’d much rather have a home or back button than a camera one (which can be mapped to the action button or, ideally, multi tapping the volume button anyway).

Except its not complete, right? There are contradictions that need solving.

I love that twist.

You ever play Transistor, by chance? You should at least watch YouTube clips, if only for its awesome soundtrack. Minus time travel, the vibe feels similar.

While less similar, I sometimes think about a similar scenario in a transcendent future, specifically Orion’s Arm. There’s no time travel there either (general relatively rules), but basically, two deep soulmates diverge. One transcends to higher levels of consciousness where, eventually, their substrate spans and encloses multiple stars, while the other person stubbornly stays the same. And I think out little vignettes, like conflicts of trust when one parter can literally simulate reality for the other, despair, worry, scenarios where even the transcendent human turns to their partner in a crisis.

Avatar fics!

I just finished and posted my first long-form story ever, a kinda maudlin, meandering, and neurotically written Korra AU. It’s a ‘what if her training was brutal and abusive,’ exploring brushed over canon concepts as things go way off canon’s track. It’s also an outlet for runaway lore in my head (like lavabending on metal, or connections between bits of spiritual lore).

The shows (and books) hit me in a grim and delicate period of my life, especially the Korra/Zuko Alone episodes. I guess they stuck with me; writing them back out helped me process some things in the real world. I daydream original settings, sure, but I feel like the structure of a franchise helps.

Hence, unfortunately, the itch is not scratched. So I’m writing out a second fic, this time set in canon, expanding past Korra’s era and into a AU of Seven Havens. But this time, it’s actually planned a little, and written in… well, slightly less fanficy prose.

What about SNW?

The vibe I’m getting is “we’re eager and optimistic, but also, things get bad, the larger landscape is kinda bad and we are trying to hold straight faces?”

It feels very 2020s.

The original Oblivion was the first AAA game I played!

IMO giving good games a second wave of hype like that isn’t a bad thing, as long as it’s not totally butchered.

…But Skyrim (specifically the latest re-release) has aged better than Oblivion, I’d say. It’s kinda past a threshold where the visuals are okay enough, and gameplay ergonomic enough, where even the original isn’t so jarring compared to, say, KCD II and all its hyper fanciness. Like, Daggerfall -> Morrowind was a different world, Morrowind -> Oblivion was like you’re in reality, Oblivion -> Skyrim was dramatic, but… stuff after that feels more like icing?

I mean, I wish TNG was a little less silly too.

That’s what I’m saying, there is no retooling. Some of AMD’s existing OEMs are already making W7900s.

Here’s the bulk of the process on the OEM side, other than maybe leaving an ECC chip off:

• Take finished W7900.

• Change ID in firmware (so the CAD drivers don’t recognize it)

• Apply a different sticker, put it in a different box

• Do the paperwork of making a new SKU, like they make for overclocked cards

That’s not that expensive. If it doesn’t sell a lot, well, not much skin off thier back. And it would make AMD boatloads by seeding development for their server cards (which the workstation cards to not do because they are utterly pointless at those prices).

This is all kind of a moot point though, as the 7900 series is basically sunsetted, and AMD doesn’t have a 384 bit consumer card anymore (nor a GDDR7 one to use the new, huge GDDR7 ICs).

Yes, it is:

https://www.amd.com/en/products/graphics/workstations/radeon-pro/w7900.html

https://dramexchange.com/

16gb GDDR6 ICs are averaging $10 each. The clamshell PCB is already made. So the cost of doubling up the VRAM in a clamshell configuration 7900 XTX (like the W7900) is like $100 at most, on top of this being a seperate memory supply from HMB the datacenter accelerators use. But AMD literally tells its OEMs they are not allowed to sell such clamshell configs of their cards, like they have in the past.

The ostensible business reason is to justify the actual ‘workstation’ cards, which are laughing stocks in that space at those prices.

Hence, AMD is left scratching their heads wondering why no one is developing for the MI325X when devs have literally zero incentive to buy AMD cards to test on.

So if AMD makes a bunch of “AI Accelerators” and nobody buys them because they would rather nVidia (which the video talked about)?

Well, seeing how backordered the Strix Halo Framework Desktop is (even with its relatively mediocre performance), I think this isn’t a big concern.

There is a huge market dying to get out from under Nvidia here. AMD is barely starting to address this with a 32GB model of the 9000 series, but it’s too little too late. That’s not really worth the trouble over a 4090 or 5090, but that calcus changes if the config could be 48GB on a single card like the 7900.

I dunno what your talking, but all AMD has to is this:

• Pick up the phone.

• Tell their OEMs VRAM restrictions are lifted.

• Put it down.

…That’s it.

They’d make seperate SKUs with double the VRAM. AMD doesn’t have to waste a cent.

one viral AI avatar or “Barbie Box” image can consume enough energy to fully charge an electric car several times.

A Model 3 battery is 200,000-300,000 kiloJoules.

Absolute worst case for an image, even taking very extreme estimates and amortizing out all the training, is like 30 kJ. Maybe 70 kJ for a slop video that takes under a minute to render, which is on the order of browsing Lemmy on a laptop for a bit. For reference, a local generation with FLUX dev on my 3090 is 2 kJ per image, and that’s relatively inefficient.

I’m just saying, that is a bad comparison, as EVs take an absolute truckload of electricity to run.

The pricing for memory is still pretty bad. $4K for 96GB, $5.6K for 256GB, $10K for 512GB. One can get 128GB on the M4 Max for $3.5K, at the cost of a narrower bus so it’s even slower, but generally, EPYC + a 3090 or 4090 makes a lot more sense.

SOTA quantization for these are mostly DIY. There aren’t many MLX DWQs or trellis-quantized GGUFs floating around.

But if you want to finetune or tinker instead of just run, you’re at an enormous disadvantage there. AMD’s Strix Halo boards are way more compatible, but not standalone yet and kinda rare at this point.