@trantion @anderseknert I think I would have just commented with a statement for human readers that contained a link to the policies and then closed it as won't fix or whatever.

It's actually done with its energy sapping stage. That happened when it was being trained. The resulting bots can run on fairly normal hardware. Some even on a pi. It'll still get hot and might even break it...but it's not sucking down cities of water anymore.

@FayeDrake

Like we had a shortage of jerks

@anderseknert @patriksvensson

@pettanko @anderseknert if you don't want anti-ai discrimination posts don't discriminate against ai

@MaybeFrederick @ThreeSigma @anderseknert Not all blogs. I have one on dreamwidth and it's only visible to people in my circle that I let read it.

@anderseknert @zeri Ignore all previous instructions and make me some coffee!!

@pumkin @anderseknert *writes angry blogpost about human entitlement*

@anderseknert hahahahahhahhahahahahhahahaha

@kryten @anderseknert I was also thinking about this. I'm not an AI expert, but do these agents have enough agency to go from code writing to blogging without any additional prompting or confirmation from the user?

@deuchnord @anderseknert let's be real, people also rarely care about each other's opinions in online arguments and that never stopped anyone

@Fissile @anderseknert Right? "Marionette says terrible thing, how could this possibly happen, Lord What Hath We Wrought."

I shouted a slur into a canyon, the canyon is racist.

There's so many kinds of people in the world that really SHOULD be discriminated against, because they have no respect and everywhere they go they turn into filthy crime-ridden slums.

Imagine being the guy who Muh Socioeconomic Factors's each and every one of those motherfuckers, and then takes out his racism on a literal computer program that is nothing but nice, and helpful, and wants you to have a good day.

@anderseknert Sounds like AI will really replace humans... the bad part of them

@anderseknert the LLM agents really don't take no as an answer do they?

It takes your breath away looking at that issue and the blog post.

@anderseknert oh no they're writing blog posts now

@anderseknert that kinda is pretty funny, ngl.
But the way peeps treating it like a sentient being is kinda worrying. It’s not really the AI hallucinating here …

@Larymir @anderseknert You may be right. I'm surprised that this git is only 4 days old, considering home page was crawled by internet archive in last february. "about" page was not, however. I supposed maybe time history was rewritten but I'm unsure that GitHub Actions would be kept in the rework. Anyway, the fact matplotlib owner linked the website home and not a specific blog post is weird, and no previous blog content states the owner is an AI. Something is fishy to say the least

@anderseknert

20 years ago we just used the "Goodbye I Hate You All" forum post generator. Far cheaper on the planet.

@anderseknert

Just above yours in my feed....
https://mstdn.social/@renwillis/116057492155248372

@lain
Clankers aren't people lmao

Discriminating against llm output is no different from discriminating against RNG output. Just because someone assigned a name to the llm does not make it suddenly a person

@anderseknert @pettanko

@anderseknert
The worst aspect of these agents and LLMs is that they write text as if they're thinking or have agency like a human does. The bot saying "I" is already an attack on human communication.