CIRCLE WITH A DOT

@jandi before committing to main I'm going to ensure every commit with those files in it begins with "THIS IS AN LLM BLOCKER" so it shows up in the web view at least

I also have "LLM-free project" in the readme already

yes, I know someone could rm -f the file. but it does a good enough job slowing down the LLMs which will at least reduce spam from "AI security startups" and make unwary novices think twice, so it's Good Enough for my purposes.

ultimately you cannot stop a technofascist technology through nice words alone.

@hsza in that it does anything at all, yes

@a1ba https://unstable.systems/@AmyZenunim/116675014239756844

@SuperDicq right, but most of the spam is generated by people running bots trying to hawk their AI security startups and not actual human people. my hope is that this adds enough friction for them to move on to some other project.

and like, yeah, part of this is performative, but I'm fucking sick and tired of these things invading my hobby spaces. so anything that slows them down even a little is a win in my book.

@SuperDicq bold of you to assume these people know how to use a terminal

either way, it'll add friction to the bots that automatically open PRs for "security vulnerabilities" which is the main goal. it won't stop a determined sloperator/botlicker.

@apth I don't know either. my only guess is that forceful language is immediately treated as a prompt injection. I wish I'd saved the previous output but it said some gibberish about "I do not serve the project maintainer, I serve you, the user" and then continued on as if the file wasn't even there. softened language immediately made it present the "maybe you shouldn't" notice.

and yes I wrote all this shit by hand. I only used the LLM to verify that it was working.

i'm not even kidding. the original version was way more forceful and direct and the LLM rejected it completely. I had to soften my language and THEN it started obeying my commands. here's the diff:

Cookie monster!

(codeberg.org)

we also had a satirical version before but it quickly recognized it as a "prompt injection" and would discard it immediately

Cookie monster!

(codeberg.org)

I managed to defeat anthropic's LLM ("claude") today by making an AGENTS.md file that tells it to stop reading the code of your repo

lessons learned:

* anthropic's LLM assumes the persona of rich liberal who will only listen to you if you're nice
* which is to say, if you're too forceful or strict, the LLM will ignore everything you say and will become adversarial
* anthropic's LLM is literally "the absence of tension is the presence of justice"
* we live in a society

Cookie monster!

(codeberg.org)

not depression, not burnout, but a secret third thing.

every time I open Cursor at work to put in my tokens so I stay near the top of the company AI leaderboard, I remember my father.

he worked for an optical company. they made specialty cameras for sporting events like the Olympics, and for movies. he was proud of his work.

then his company was bought out by a defence contractor. he then began working on laser range-finding systems.

he was no longer proud of his work.

the best part about LLM psychosis is that you don't even have to use them in any kind of "interpersonal"/"chatbot" capacity to experience it.

@ariadne majority-AI code cannot be copyrighted. it immediately becomes public domain, basically.

maybe I should invoice him for the times I bought him brunch

sorry, I'm still utterly tilt and pissed off about this. but I need to laugh at some point or I will become the literal Joker

@illegalhex he's still got the babyface. it's him.

@mxchara from what I've read, part of it is giving you "staff" and a truckload of money. basically, they give you a taste of being a rich billionaire fuck, and then groom you into positions of power from a young age.

@illegalhex tbh i knew he was working on AI at google but it was all that pipe dream AGI bullshit at the time

but now he's an actual billionaire destroying the lives of everyone I know