If LLMs can’t be made more power efficient, then the technology isn’t ready yet for general use. There may have to be fundamental engineering breakthroughs before what these systems do can be more than a very expensive and inefficient worldwide proof of concept.

Of course, the military and the surveillance industry will still want to keep it, even if no one else can afford it, and we’ll be the ones paying for it.

But without widespread usage there won’t be any breakthroughs… Anti-LLM folk could redirect their attention to the core problem (energy source, wind and solar shouldn’t raise any concerns) but they prefer to attack the LLMs themselves. Early implementations of every technology were crazy energy ineffective at the moment when becoming practically useful.

No one should be arguing if LLMs are useful or worth investing into, they are and it can be a really useful tool, but the rich (seemingly spinless) lizards have managed to skyrocket the price of RAM, pushed global warming to new extrimes and jacked up electricity prices. Oh and while making an AI bubble so big and fradgile, that when (not if) popped will send us back to the great depression! All in the percute of “the next big thing”.

to;dr LLMs are a cool tech but in the percute of it have managed to fuck up a whole bunch

pursuit*

LLMs in the context of capitalism will always be used against the working class.

Anti-LLM folk don’t just have an issue with energy consumption. There’s also the issue of intellectual theft, studies that correlate worsened education with LLM usage, and the tech leadership silencing those issues by corrupting politicians. But I’m sure if they get past that they’ll understand how great LLMs are.

We understand LLM is great at something but we can never get passed the facts that you’ve stated and it also isn’t great in all thing like we’re being sold to and forced to look at everyday. If the tech stay with the professional and not given to the general public, if there’s no cunts telling everyone to adapt or get thrown out, if there’s no mass disruption toward people’s living, i’m sure it would be much more acceptable. The biggest issue with the tech is the cunt parading it.

Oh no, we won’t get breakthroughs in slop.

Whatever will we do?

LLMs kill kids

If LLMs can’t be made more power efficient, then the technology isn’t ready yet for general use.

Hm, I don’t see the harm in powering local models with solar.

Depends. Is it solar energy that started production to power these models? Great! If not, then you’re still consuming power and the difference will have to be made somewhere, likely by burning more fossil fuels.

Oh yeah, I mean my own personal solar modules in my backyard. I agree it’s not that easy to say “solar exists, checkmate doubters”, that would be silly.

OK, that is fine. I’ve seen a lot of people saying stuff like “Me mining crypto doesn’t matter because I live next to a wind farm” so I’m a bit wary.

Well that is an alternative. But I bet each query to a local 27b model takes less energy anyway than the same query to Claude Opus or Fable in some data centre. So it is also a case of improving energy efficiency by running smaller models.

They CAN be made more efficient. Hell, AMD just bought a company that translated LLMs to direct silicon... which should enable scalability if it wasn't for one of the key players owning the tech thus limiting its use.

It's also estimated that some 70% of Claude tokens gets spent on recursive self-correction so the output is higher quality. Instead of bettering the model, Anthropic just keeps throwing money at it to make the problem go away.

Meanwhile Chinese models - as small as 36-40-80b params - are outdoing these fancy frontier models if the right harness is used.

I've spent the last month or so using nothing but a local 27B Qwen model as my coding assistant and it's working fine. The big American AI companies could all collapse in ruins tomorrow and it's not going to slow things down all that much, anyone who wants to use AI will be able to continue using it. Might even be better in the long run if that happens - the people who are scared of ASI will get their much-demanded temporary slowdown at the frontier while everyone else catches up with open-weight models that everyone can use without central control.

Pity about the economy in the meantime, I guess.

I’m always running into context limit problems

Wanted to refactor a too large and hectic hacked together state machine into separate functions and reduce redundancy and increase code readability

Every model I’ve tried failed.

And that should be an easy task an actual AI should be able to handle. Logic is done, just translate to a different structure.
That should be exactly what they are able to do at least.

None worked, because the project is just to big and interlinked.

I've got 24GB of VRAM, which helps a lot. I'm using the Q4_K_M version of this model, with the KV cache quantized to Q8, and that's able to get me to 128000 tokens of context. I'm using the Qwen Code harness in VS Code.

If 128000 tokens still isn't enough, break the task down into smaller chunks and help the harness out with MCP tools - I installed Codebase Memory to help it navigate large codebases more efficiently, for example. Ask the coding agent what's causing problems and it may be able to help identify solutions too, I've had it troubleshoot itself several times now and it's been quite good at inspecting its own settings and finding where the problems are with them. I even installed Firecrawl for it so it can do its own web searches.

anyone who wants to use AI will be able to continue using it

Anyone who’s privileged enough to have the hardware.

Anyone who lacks the skill to to it themselves…

Open models will still be available via API from third-party hosts. The bigger ones like Kimi K3 or GLM 5.3 don't fit on conventional desktop computers to begin with, most non-institutional users of those would already be doing that (eg via OpenRouter).

The kind of specs that would enable local LLM usage were available on laptops since 2020. The floor for acceptable performance for LLMs stopped literally just before 24+gb of VRAM would have become pretty common in consumer applications. The reason this technology isn’t more ubiquitously available is because ‘they’ thought it would be more profitable to hoover up literally every byte of RAM and GPU in the world rather than let the technology continue to naturally evolve to a place where any consumer could have run useful models on their home computers. In other words, your anger is misplaced.

In other words, your anger is misplaced.

What’s your assumption about where my anger is directed?

local 27B Qwen model

How much VRAM does that require?

What kind of hardware do you need to run a model like that? And what do you use :-) ?

I run a 5 bit quantized Qwen 3.8 27b at home on a gaming PC with an RX 7900 XTX and RAG memory using Gemma MoE (i forget what one)running on an old 1080 ti on my Valheim server.

It takes 3-15 minutes to run depending on how many tool calls the model makes and that’s mostly memory or web scrapes since that runs on the slower hardware. I mostly just use it as a discord bot to look stuff up for me and to fetch the news.

Sabata, goddess of computing, can I run it on a 32GB system with a 4GB nVidia GPU?

Probably not, or if it runs it may take hours or days to get a full output. VRAM on the GPU really decides what will run reasonably. Try an 8 billion parameter phone sized model and see if its faster than sending a letter, then keep moving up until your setup can’t handle it.

8B models are a bit too dumb to use any tools reliably but should handle a basic conversation.
llama cpp will allow you to manually load a model and tweak settings.

unsloth will give you a basic setup with some tools, manage model downloads, and settings.

If you’re talking about the company I think you are, that acquisition by AMD is interesting news.

The company was the one with a demo AI in their page that was super quick??

I always figured that was the future of AI when I saw it. Where you’d buy processors with n number of “AI cores” that’d have work/prompts scheduled as needed in the same way we now CPUs and RAM.

Is anyone trying to make LLMs with analog computing + memristors? That would make them absurdly efficient

I read the Wikipedia page for memristor but didn’t understand it. What’s it do?

I was working on an abacus-based LLM, but I ran out of beads.

Hello friend, I am building a new bead factory in Taiwan.

Just give lend me your money and I will give you beads in the future sometime.

lol, I wrote a reply about a company etching onto silicon and then go down and saw yours. Oops.

I hadn’t heard AMD bought them though. Looks like they’re working on a 20b model as well for this year.

There may have to be fundamental engineering breakthroughs before what these systems do can be more than a very expensive and inefficient worldwide proof of concept.

Controversial thought after reading this sentence: It’s cryptocoins all over again.

That'd be fine. Cryptocurrency as a whole is still going strong (total market cap is around 3 trillion dollars) and most of the major ones have switched over to energy-free proof-of-stake validation.

Cryptocoins deliberately burn more energy doing useless calculations to prevent inflation. That’s a little bit of an oversimplification, but I don’t feel like explaining the full thing and the vibes are there.

There’s this one company that has etched a small model onto the silicon directly, and can get something like 15k TPS from it at very low electricity usage, and is working in production.

They claim, they can go from model -> chip in a very short time (months not years), and that at datacentre scale, it would pay for itself in about a year vs todays infrastructure, but they have yet to prove they can properly etch the bigger models across multiple chips and have it all work together like the small model.

The downside if it works at that scale is the model will be stuck on whatever you etch into it. They did say you can apply Loras ontop of it though, so you can tweak it once built.