vivendi @ vivendi @programming.dev

Posts

0

Comments

138

Joined

3 mo. ago

I don't fucking understand why Lemmy is permanently stuck in 2023 with AI

Using RAG (retrieval augmented generation) results in much lower, almost negligible confabulation rates

Guys PLEASE PLEASE PLEASE, download it all. The only way to preserve information is to copy it until 1 survives

I am just a poor fucking Iranian with shitass internet and no money to buy a NAS but I'll try to hawk some part of it as much as I can

Based and GNU pilled

This is why I license my work under GPLv3+ (not going to link my codeberg/GH because I'm not fucking stupid)

download "LM Studio" and you can download models and run them through it

I recommend something like an older Mistral model (FOSS model) for beginners, then move on to Mistral Small 24B, QwQ 32B and the likes

Stalemate means Ukraine is fucked because a war of attrition potentially without US support doesn't look good for them at all

Like, it's amazing they managed to hold shit together for all this time and all, but damn they're looking pretty fucked ngl

First, please answer, do you want everything FOSS or are you OK with a little bit of proprietary code because we can do both

Basically? Wartime propaganda

Ukraine has been doing individual, small wins like this and they obviously toot their horn when it happens

But on a large scale, Ukraine has been slowly losing ground

Fuck ClosedAI

I want everyone here to download an inference engine (use llama.cpp) and get on open source and open data AI RIGHT NOW!

OK you walk the walk I respect that

If you could be arsed, it's entirely possible to train an LLM with 10 PCs using strictly open source and ethical datasets available on huggingface (although not something that is going to beat SOTA models)

That is a one time cost. Do you add the cost of every single watt of electricity ever generated, the cost of making factories, extracting materials and so on, to the "climate impact" of electrical rail service or EVs?

This is just hysteria, I'm going to be absolutely deadass with you

Also, you theoretically could make an LLM entirely on your own PC, it'd just take a bit of time

This is .... Incredibly dumb.

The whole climate impact of AI is overstated a million times over. You can run a perfectly capable, GPT 4o destroying LLM on your own GPU right now. What is the difference between jonkling to vidya gaems or running an LLM for doing something productive for a fucking change?

The servers running the network you posted this shit over , the servers used to develop your favorite Gacha coom game, the servers used to deliver your shit ass Xitter feed, or your Netflix garbage, use more power than AI and LLMs are actually actively more productive than that shit

SCMP is one of the most libshit news in China

Also, chill it cracker, it's news about a reactor, we don't need your state dept. programming here

Another banger from lemmites

Mate, you can use AI for porn

If literally -nothing- else can convince you, just the fact that it's an automated goon machine should tell you that we are not going to live this one down as easily as shit like NFTs

Not with alcohol, just in general

Also you'd be surprised at how many "muslims" drink and fuck and gamble and do any manner of debauchery possible

Funny. We Iranians almost always eat raw onions alongside food, but everyone in the west seems to hate them unless it's dripping with 6 liters of frying oil

It is the united states, people already be living like that before the nukes

I designed microcontrollers and wrote assembly for them

Now I'm just a regular software dudebro

What class do I get?

Holy fucking shit

How did you find this archeological piece?

God I hate this. I much prefer writing to videos.