LMFAO

For sure, I think they genuinely think if AI got good enough they could just cull all the pesky workers at that point, and live like gods.

For sure, they’ve probably dropped more significant papers in the past year than any other groups. It does seem like the mindset in China is very different overall though. In the states, it’s basically a cult at this point where they’re trying to build a god with AGI. In China, it’s just treated like another tool for automation and companies see it as common infrastructure, akin to Linux, that people will build interesting things on. Hence why pretty much all the models in China re developed on open basis. Everybody there seems to realize that there’s no real path towards monetizing the models themselves.

It looks like you can run a low quant version on a 125gb machine, and apparently performance is still really good. https://github.com/makepad/llama_antirez_deepseek

Same

I’m expecting we’ll see the same dynamic we saw with solar panels and EVs. Once Chinese companies ramp up production, there’s going to be a flood of affordable GPUs. The big question is whether Western countries will ban them for ‘national security’ reasons.

Roghlt 10 to 13 tps give or take.

🤣

I do love how steam engines remain peak energy production technology that humanity has managed to come up with.

I’m using 3.6 at 27bln and q8 on a M1 with 64gb.

I think so yeah, searxng is definitely the most privacy focused option.

I’ve just been using the builtin, but searxng might be better. Seems like a lot of people prefer it.

I’ve stopped bothering using an editor with LLMs. I just get the model to make a phased plan, write using TDD, and tell it to do staged commits for each feature. Then I just review the diffs after.

I ended up settling on opencode, but I find all of them work more or less the same nowadays. Pi is an interesting one which is very minimalist.

I haven’t tried comparing them myself, I guess you just kind of have to gauge if it works well enough. :)

It’s honestly incredible how good the local stack is nowadays. It’s literally better than any frontier model you could’ve rented like a year ago.

It’s entirely possible we might see fairly capable models that can be run with 16 gigs of RAM in the near future. Qwen 3.5 came out in February, and you needed a server with hundreds of gigs of memory to run a 397bln param model. Fast forward to a couple of weeks ago and 3.6 comes out with a 27bln param version beating the old 397bln param one in every way. Just stop and think about how phenomenal that is https://qwen.ai/blog?id=qwen3.6-27b

So, it’s entirely possible people will find ways to optimize this stuff even further this year or the next, and we’ll get an even smaller model that’s more capable.

I’m sure it’s applicable outside AI as well, but I guess they’re developing these chips to run and train models.

Mainly data sovereignty. Running a local model means all your data stays on your machine. Any time you use a service you’re sending whatever the model is working on to the company. Another advantage is the price. With services you have to pay a subscription, with local models you get to run them for the price of electricity.

16gb is a bit low unfortunately. You could run a 2 bit quant of latest Qwen, but that’s going to be a severely degraded performance. https://huggingface.co/unsloth/Qwen3.6-35B-A3B-GGUF

Might be worth trying though to see if it does what you need.

the article literally links to how BYD is putting this stuff into production https://interestingengineering.com/energy/catl-ev-battery-7-minute-charge

https://www.intelmarketresearch.com/ion-exchange-membranes-for-vanadium-redox-flow-battery-market-41675

Weird how Americans have such poor reading comprehension. Oh wait, I guess it’s your education system.

why?

seems like where things are headed

it seems like the whole drama didn’t really impact usage of their models by government agencies though

• https://www.axios.com/2026/04/19/nsa-anthropic-mythos-pentagon
• https://www.politico.com/news/2026/04/14/anthropic-mythos-federal-agency-testing-00872439

I’m still convinced this was just pure theatre to present them as the good and independent company that’s totally not working with the fascist state.

Yeah, I expect it’s going to be rolled out a lot more sanely in China than it will be in the west.

I mean if I was paralyzed I’d take my chances with the chip.

There’s plenty to understand. Synthetic apertures allow getting a far higher resolution than you could otherwise which mean that you can place satellites in higher orbit to get the same coverage you’d get with more satellites in a lower orbit. This is what the article says, but you clearly have no clue regrading the subject and just need to argue for the sake of arguing. Go touch some grass.

Sounds like you need to read up on what synthetic aperture radar is to understand the article.

wow what an amazing counterpoint you got there

I absolutely hate the fact that every chat app now has a proprietary protocol and you have to install a bunch of apps to talk to different people. It’s such a travesty.

Ah didn’t know about those, I only knew of electroplans.

true, but only as prototypes that never went into mass production

Perhaps they mean commercially produces ones. To my knowledge these have never gone past experimental prototype stages.

Ground effect vehicles are basically airplanes that are forced to fly really, really low. They take off from water and cruise just a few meters above the surface. At that altitude, the air gets compressed between the wing and the ground or water, which creates a huge cushion of extra lift. This lets the vehicle carry way more weight than a normal plane of the same size and power, making it incredibly efficient for hauling cargo over water. The trick is that it only works over flat surfaces like oceans or lakes, and the piloting can be tricky because you’re skimming the waves at high speed without actually being able to climb to a higher altitude. It’s a neat piece of engineering that trades operational flexibility for raw lifting power.