refusing to use LLMs: not purity culture

@0xabad1dea this needed saying. I think it is going to continue to need saying for a while unfortunately

@mjdxp @0xabad1dea "we broke a good thing and then made a worse thing that's only ~maybe~ better than the thing we left broken because we broke it" really sucks and it's *so much of tech* right now

@0xabad1dea @Razemix cory doctorow wrote in his anniversary blog post that disowning LLMs as a technology is similar to "neoliberal purity culture"

@mjdxp @0xabad1dea SEO garbage is a real problem, but I don't think something that is capable of making up plausible sounding text, that is disconnected from reality, is the best solution.

I'd much prefer something that only ever surfaced text that has a known, live source I can check, both to be sure it is actually there, and to check the surrounding context.

Feels like LLMs are crowding out better solutions to some problems.

@morganist @0xabad1dea Oh, wow… Thanks for the context.

@0xabad1dea if yelling at idiots who legitimize slop machines is purity culture so be it i won’t stop doing it

@quantumsys @0xabad1dea what word should i use that’s not ableist or saneist? to call slop advocates fash wouldn’t be quite right

and you’d need to clarify how exactly “shaming working class consumers” relates to my point

@quantumsys @0xabad1dea look sometimes beings are too far gone for me to bother engaging in good faith and arguing the same points over and over again when i know it won’t go anywhere. my energy for doing that is finite. so in such cases my response to [insert non-saneist word here]s doing apologism for fash is going to lack any kindness, and i think that’s quite reasonable. when i know i can’t inspire change in them, i can at least yell a little to give myself and those on my side of the discourse an affirmation.

@quantumsys @0xabad1dea

OP feels like it’s implying a general claim that calling people out on using unethical tech is wrong, which is a bad and annoying albeit frequent take, might’ve misinterpreted that

@0xabad1dea

> one missed coffee from doing a terrorism

I know this was intended as a bad thing but mind of want this on a t-shirt (in jest)

@0xabad1dea@infosec.exchange I don't dehumanize people but I wanna do a terrorism on AI datacenters. ​​

100%

Also. There's clearly a distinction between the AI booster that uses LLM on their on volition and promotes their use and the worker that see themself forced to use them by their employers.

I've seen people getting insulted and shouted at when they had made it pretty clear that they were been pushed to use them by the companies they worked for and they didn't even supported their use.
Shouting at a worker because they are forced to use technology made to exploit them is not a good look.

@0xabad1dea i used an llm once to try to create a name for something. i ended up scrapping the thing that needed the name though. other than that, mostly to figure out how to trick them. at this point i refuse to send a single prompt to one if i can help it. (i didn't Know yet.)

@morganist could you give a link?