The bot isn’t really cognizant of its own replies.

…Well, it could be with a postfilter, ranging from sanity checking its own reply, to a tiny separate safety model (of which many exist), to a basic keyword check. But apparently Twitter has no resources for such things.

But the point is this isn’t really the intended design.

I’ve used 5K some.

IMO the only ostensible benefit is for computer type stuff. It gives them more headroom to upscale content well, to avoid anti aliasing or blurry, scaled UI rendering, stuff like that. 4:1 rendering (to save power) would be quite viable too.

Another example would be editing workflows, for 1:1 pixel mapping of content while leaving plenty of room for the UI.

But for native content? Like movies?

Pointless, unless you are ridiculously close to a huge display, even if your vision is 20/20. And it’s too expensive to be worth it: I’d rather that money go into other technical aspects, easily.

And that’s why conservative leaders hate Wikipedia.

It organizes and filters information. Can’t have that.

Are you sure about that? You likely use DPI scaling at 4K, and you’re likely limited by physical screen size unless you already use a 50” TV (which is equivalent to 4x standard 25” 1080p monitors).

8K would only help at like 65”+, which is kinda crazy for a monitor on a desk… Awesome if you can swing it, but most can’t.

I tangentially agree though. PCs can use “extra” resolution for various things like upscaling, better text rendering and such rather easily.

I respectfully disagree. Folk’s eyes are ‘used’ to 24P, but native 48 or 60 looks infinitely better, especially when stuff is filmed/produced with that in mind.

But at a bare minimum, baseline TVs should at least eliminate jitter with 24P content by default, and offer better motion clarity by moving on from LCDs, using black frame insertion or whatever.

(Most) TVs still have a long way to go with color space and brightness. AKA HDR. Not to speak of more sane color/calibration standards to make the picture more consistent, and higher ‘standard’ framerates than 24FPS.

But yeah, 8K… I dunno about that. Seems like a massive waste. And I am a pixel peeper.

It’s all about the baseline.

Cinematic, Blu Ray bitrate 1080p vs 4K is not too dramatic.

Compressed streams though? Or worse production quality? 4K raises the baseline dramatically. It’s much harder to stream bad-looking 4K than it is 1080p, especially since ‘4K’ usually implies certain codecs/standards.

Ashamed but not ashamed to admit it, but I have quite a fascination with this, and am working/thinking on a fanfic that explores Avatar lore more scientifically. A major perspective is through a Zaofu engineer-type character who looks at spirits, his own bending, and other lore like the portals or vines more scientifically.

That’s where my brain goes. Shrug.

I was hoping we’d get a “bender engineer” character in canon, but that doesn’t appear to be the direction they’re going.

And yet the article is basically all upvotes.

As of late, Lemmy has been feeling way too much like Reddit to me, where clickbait trends hard as long as it affirms the environment.

I’ve even pointed this out once, and had OP basically respond with “I don’t care if it’s misinformation. I agree with the sentiment.” And mods did nothing.

That’s called disinformation.

Not that information hygiene is a priority here :(

Yeah, comments often “correct” that, but that doesn’t stop the extra order of magnitude of exposure the original post gets.

As much as the Twitter format sucks, Lemmy could really use a similar “community note” blurb right below headlines.

Preface: I love The Guardian, and fuck Altman.

But this is a bad headline.

Correlation is not causation. It’s disturbing that OpenAI even possesses, and has mined for these statistics, or that millions of people somehow think their ChatGPT app has any semblance of privacy, but I’m reading that millions reached out to ChatGPT with suicidal ideations.

Not that it’s the root cause.

The headline is that the mental health of the world sucks, not that ChatGPT inflamed the crisis all of the sudden. The Guardian should be ashamed of shoehorning in some “Fuck AI” article into that for clicks, when there are literally a million other malicious bits of OpenAI they could cover. This a sad story, sourced from an app that has an unprecedented (and disturbing) window into folks psyche en masse, they’ve twisted into clickbait.

Doctorate material on sub-elements:

• Metallurgy and crystallography (metalbending).

• Volcanism (lavabending).

• Circulatory systems (bloodbending).

• Electrostatics and plasma physics (lightingbending).

• Quantum mechanics (spirtbending and astral projection).

For airbending? Yeah, atmospheric science and fluid dynamics is probably the better manga.

You don’t have to overclock, but I’d at least look at your mobo’s settings. Many mobos set really, really bad, non stock settings by default, especially with XMP memory.

An example: they might default to 1.3V VSOC which is absolutely a “default overclock” and is going to make idle power skyrocket, and your CPU potentially unstable because infinity fabric doesn’t like that. For reference, I personally wouldn’t go over 1.2V VSOC myself and shoot for like 1.1V.

I’d recommend Buildzoid’s videos:

https://youtu.be/dlYxmRcdLVw

https://youtu.be/Xcn_nvWGj7U

And Igor’s Lab for general text info.

Also, if you don’t turn on XMP at least (aka the RAM’s rated speed), they will run at some slow default and hurt your inference speed rather significantly.

Depends on the culture.

The Air Nomads were segregated in Avatar, but they were sexually progressive forever, and raised children communally. Stuff like that is fine.

I know nothing of Smurf lore, but they don’t seem like one of those repressive segregated cultures.

For DDR5? Depends how much you care about latency:

https://pcpartpicker.com/products/memory/#ff=ddr5&Z=131072002&B=1000000000%2C1250000000&sort=price&page=1

The $342 Crucial kit is kinda a no-brainer for 128. Its timings aren’t great when overclocked, but it’s 5600 MHz out of the box, low voltage, and significantly cheaper per gigabyte than many 64GB/96GB kits. See for yourself:

https://www.igorslab.de/en/when-size-is-all-that-matters-crucial-2x-64-gb-ddr5-5600-kit-with-new-micron-32-gbit-ics-in-test-incl-gaming-and-overclocking/4/

The overclockability matters even less if you are on a 7000 series CPU.

I got the 1.25V Flare X5 kit because I wanted tighter timings for sim games, albeit at a MUCH lower price ($390) than it is now.

RAM prices seem to be rising (hence the price of my kit spiked by $200), so now is not a bad time to buy.

To what?

64G would be good, as that’s enough to fit GLM Air. There are some good 2x64GB 6000Mhz kits for 128GB as well.

I have access to GLM 4.6 through a service but that’s the ~350B parameter model and I’m pretty sure that’s not what you’re running at home.

It is. I’m running this model, with hybrid CPU+GPU inference, specifically: https://huggingface.co/Downtown-Case/GLM-4.6-128GB-RAM-IK-GGUF

You can likely run GLM Air on your 3060 desktop if you have 48GB+ RAM, or a smaller MoE easily. Heck. I’ll make a quant just for you, if you want.

Depending on the use case, I’d recommend ERNIE 4.5 21B (or 28B for vision) on your Macbook, or a Qwen 30B variant. Look for DWQ MLX quants, specifically: https://huggingface.co/models?sort=modified&search=dwq

Not anymore.

I can run GLM 4.6 on a Ryzen/single RTX 3090 desktop at 7 tokens/s, and it blows lesser API models away. I can run 14-49Bs (or GLM Air) in more utilitarian cases that do just fine.

And I can reach for free/dirt cheap APIs called locally when needed.

But again, it’s all ‘special interest tinkerer’ tier. You can’t do that with ollama run, you have to mess with exotic libraries and tweaked setups and RAG chains to squeeze out that kind of performance. But all that getting simplified is inevitable.

Anything even tangentially AI, anything that even sounds like AI gets mass downvotes on Lemmy.

Don’t get me wrong, Sam Altman is an even bigger con artist than Musk. But outside the self hosting niches, Lemmy skews towards whatever the opposite extreme of ‘tech bro’ is.

Net might get to where you need AI

I hate to say it, but we’re basically there, and AI doesn’t help a ton. If the net is slop and trash, there’s not a lot it can do.

Hopefully by then they will have figured out a way to make it free.

Fortunately self hosting is 100% taking off. Getting a (free) local agent to sift through the net’s sludge will be about as easy as tweaking Firefox before long.

You can already do it. I already do it (and am happy to ramble about how when asked), but it’s more of an enthusiast/tinkerer thing now.