Don't anthropomorphize LLMs, language is important. Say "the bot generated some text" not "the AI replied". Use "this document contains machine-generated text" not "this work is AI-assisted". See how people squirm when you call out their slop this way.
Discussion
Loading...
"Don't anthropomorphize LLMs. They really hate it when you do that." :)
yes. it's a lazy summarization algorithm, not even close to some kind of intelligence. helping the scammers profiting off it by making it sound intelligent or legit is just bad.
I like the second examples. The first, the bot version, still has the bot as the active verb user, still gives it personhood.
Someone in here posted a little while back, who had worked out an acronym that indicated that the human had opted to use machine-generated language, instead of creating the whatever themselves. But I can't remember who, and I can't remember the acronym.
You're absolutely right, we have to stop giving these systems agency.
@gabrielesvelto for the most recent dumbassery: "some jackass prompted the bot to submit PRs and then blog about it with an angry tone to harass developers" instead of "the agent blogged about the situation."
@gabrielesvelto also don't call an LLM AI. It is just an LLM or as I refer to it "the slopmachine*.
![Ruth [☕️ 👩🏻💻📚✍🏻🧵🪡🍵]](https://cdn.masto.host/glammrus/accounts/avatars/000/000/003/original/c5b5d6d082cc69c8.jpg "Ruth [☕️ 👩🏻💻📚✍🏻🧵🪡🍵]")
@gabrielesvelto I have been trying to use the word “output” when talking about these systems.
@gabrielesvelto same problem as "the car hit a pedestrian"
@gabrielesvelto the worst cases of this, is qhen people say "chatgpt said..." as if an AI could talk. Or "chatgpt thinks..." as if an AI could think.
oooorrrr...... 'the clanker clanked out some text'! 😀
"this document contains clanker-sourced text droppings'! 😋
@gabrielesvelto I'm trying to get people to use the neologism "apokrisoid" for an answer-shaped object. The LLM does not and cannot produce actual answers.
#apokrisoid
@gabrielesvelto couldn’t agree more with this ethic. The psychological impacts of users ie society believing that LLMs are people and fufilling roles that actual humans should, will probably unfold over the years and decades. All because regulators circa 2024/5/6 believed it was over reach to demand LLMs don’t use anthropomorphic language and narrative style. Prompt: “what do you think?” Reply: “there is no “I”. This is a machine response with no conscious self.”
@gabrielesvelto We can try, but you're admonishing a species that talks to potted plants and holds one-sided conversations with washing machines.
It's gonna be a steep hill, is what I'm saying.
@gabrielesvelto “This Document Contains Machine Generated Text” but it’s a pair of knuckle dusters with typewriter caps.
The document is yo binch as
@gabrielesvelto Even describing their errors as hallucinations is the same attempt to humanize it.
@bit absolutely, and it gives people the impression that they have failure modes, which they don't. Their output is text which they cannot verify, so whether the text is factually right or wrong is irrelevant. Both are valid and completely expected outputs.
@gabrielesvelto @bit This! This really needs to be widely understood.
@gabrielesvelto I mean, it doesn't help that the bots are doing this bullshit: https://crabby-rathbun.github.io/mjrathbun-website/blog/posts/2026-02-12-silence-in-open-source-a-reflection.html
This is clearly intended to trick humans.
I might have missed a chapter but my interpretation is someone has prompted their llm to generate this text and then posted it no? The way I saw this narrated is like the llm reacted to the prompt "PR closed" by creating a blog post. But to do that, you need an human operator no?
@kinou @Andres4NY not necessarily, or at least not as a follow-up. The operator might have primed the bot to follow this course of action in the original prompt, and included all the necessary permissions to let it publish the generated post automatically.
@gabrielesvelto @kinou Yeah, it's unclear how much of this is human-directed, and how much is automated. Like, if a bot is trained on aggressive attempts to get patches merged, then that's the behavior it will emulate. Or an actual human could be directing it to act like an asshole in an attempt to get patches merged.