I bought a datacenter GPU that doesn't fit in a normal motherboard, macgyvered the fan with jumper wires, and now I'm running a model that ties with Claude Sonnet 4.6 on benchmarks, all for £200.
I don’t know whether you read the article. It was quite interesting to me and gave me some good ideas. I not going to replicate what he did, but he has shown me a methodology I could apply to do other things.
It’s so clearly written with the aid of an LLM that I’m embarrassed for all the people downvoting this comment. Just because the basic outline of the information presented and the sequence of events comes from the real experience of a real person, it does not mean that the real person wrote the article themselves. Seriously, how many real people write like a bad 80s detective movie script? It’s literally a parody trope at this point.
I am a monkey, I live on a rock, in the middle of an infinite empty space, that none of the smartest monkeys on this rock can figure out.
I am part of a species that has spent its entire history killing itself in the most horrifying and ghastly ways for the most benign reasons that the greatest war approaches us and everybody is so accepting.
The planet I live on is slowly becoming inhabitable to point where the thought of kids is pointless. And banks have used to many money glitches that currency across the globe is going into hyperinflation.
I really really cannot illustrate how much I do not give a fuck about the cadence of written words in that small space of relief I get from the inevitable destruction looming in a completely absurd and pointless existence.
I thought parts felt like AI and parts felt like a British human writing it.
AI(?):
But here is the thing: this is a Volta GPU with 16GB of HBM2 memory, 5120 CUDA cores, and I picked it up for about £150 on eBay. The compute is still real. The VRAM is still real. And the memory bandwidth is where it gets genuinely surprising.
Human (?):
So I shoved some jumper wires into the connector and jammed the other ends into a spare fan header (turn your volume up)
I suspect he’s used some AI polishing tool to turn his draft into a publishable post.
It has nothing to do with the subject of the article. Are you blind? It’s the exact same bland, AI regurgitated writing style as every other AI written article.
For well-known authors with a large corpus of prior writing, it’s simple to run a style match analysis, if you wish to demonstrate that they lazily used generative AI to write a particular piece.
But for all the knee-jerk “ai slop” accusations I see on almost every post now, the only justification most offer is that it’s obvious, which sounds like “I can tell by the pixels.”
I detailed several of the indications in another comment. If you’re too blind to see the very obvious indicators then oh boy. That’s what makes it AI slop. They don’t even bother trying to hide it. It’s like bad CGI, there’s no need to say “I can tell by the pixels” because it’s transparently obvious.
It’s literally AI slop. Ignoring the subject of the article (which is whatever), the article is clearly entirely written by AI. The AI writing style is horrendous. “It’s not blah. It’s blah.”, “it’s genuinely <insert adjective of amazement>”
But here is the thing: this is a Volta GPU with 16GB of HBM2 memory, 5120 CUDA cores, and I picked it up for about £150 on eBay. The compute is still real. The VRAM is still real. And the memory bandwidth is where it gets genuinely surprising.
Is nobody extremely tired of the AI slop articles??
The guy is explaining how to hack around GPU fans which were never meant to be run at lower speed through jumper cables and you complain about slop?
This “guy” isn’t explaining anything. The entire article is written by AI. It’s the same bland writing style everywhere now.
I don’t know whether you read the article. It was quite interesting to me and gave me some good ideas. I not going to replicate what he did, but he has shown me a methodology I could apply to do other things.
Who gives a shit? You ever read a legal document? They’ve had the same writing style for 70 years.
Actually if you’re interested in running a local llm it’s quite a good article
It’s so clearly written with the aid of an LLM that I’m embarrassed for all the people downvoting this comment. Just because the basic outline of the information presented and the sequence of events comes from the real experience of a real person, it does not mean that the real person wrote the article themselves. Seriously, how many real people write like a bad 80s detective movie script? It’s literally a parody trope at this point.
Honestly, I just don’t care.
I am a monkey, I live on a rock, in the middle of an infinite empty space, that none of the smartest monkeys on this rock can figure out.
I am part of a species that has spent its entire history killing itself in the most horrifying and ghastly ways for the most benign reasons that the greatest war approaches us and everybody is so accepting.
The planet I live on is slowly becoming inhabitable to point where the thought of kids is pointless. And banks have used to many money glitches that currency across the globe is going into hyperinflation.
I really really cannot illustrate how much I do not give a fuck about the cadence of written words in that small space of relief I get from the inevitable destruction looming in a completely absurd and pointless existence.
word
I thought parts felt like AI and parts felt like a British human writing it.
AI(?):
Human (?):
I suspect he’s used some AI polishing tool to turn his draft into a publishable post.
This article ISN’T AI slop.
It has nothing to do with the subject of the article. Are you blind? It’s the exact same bland, AI regurgitated writing style as every other AI written article.
For well-known authors with a large corpus of prior writing, it’s simple to run a style match analysis, if you wish to demonstrate that they lazily used generative AI to write a particular piece.
But for all the knee-jerk “ai slop” accusations I see on almost every post now, the only justification most offer is that it’s obvious, which sounds like “I can tell by the pixels.”
I detailed several of the indications in another comment. If you’re too blind to see the very obvious indicators then oh boy. That’s what makes it AI slop. They don’t even bother trying to hide it. It’s like bad CGI, there’s no need to say “I can tell by the pixels” because it’s transparently obvious.
Yes but this isnt one of them.
It’s literally AI slop. Ignoring the subject of the article (which is whatever), the article is clearly entirely written by AI. The AI writing style is horrendous. “It’s not blah. It’s blah.”, “it’s genuinely <insert adjective of amazement>”
Its not AI writing. It reads exactly like someone casually talking. Its consice and to the point. There is no “its not x, its y”
He has been blogging dor dam near 6 years you can go read his older posts the writing is consistent.
It’s AI slop. You’re blind.
So much so…