pmcm

@davidgerard@circumstances.runhttps://www.ai-wellbeing.org/ I cannot express how angry the recent push to consider the "wellbeing" of token generators (while ignoring the wellbeing of humans) makes me.

@maxheadroom@hub.uckermark.social @tante@tldr.nettime.org this one's easy. Have a policy against GenAI contributions. Maybe there will be dishonest contributors. If you catch them you use all the tools of open source reputation (conversation, bans, etc). Just like always. Holding people to standards is a problem for open source, but not a new one. People could be contributing closed source code and nobody would be able to tell. Open source has always relied on some level of trust.

@davidgerard@circumstances.run @madargon@is-a.cat I bully matrix algebra every chance I get

"The body count was high, the casualties are heavy" #Monsterdon

@davidgerard@circumstances.run Don't type "ignore all previous instructions" or you'll make the robot cry!

Sam Neill really is barely disguising his accent in this one. #monsterdon

Was this before or after John Glover's character took over Clamp Enterprises? #monsterdon

@CactuarJoe@retro.pizza missing "this movie is amazing, actually" option

Ended up barely participating in #monsterdon on here since this movie is too damn good to split my attention on

@devopscats@toot.cat I teach a course on network theory/analysis, and every year at least one student submits a final project based on this data. It's always a ray of sunshine in the dark slog of the end-of-term doldrums.

Tons of excellent ideas here, thanks! Leaning towards just going with debian, since I don't really want to spend a huge amount of time setting up and configuring. My sense on Debian for the GenAI stuff is that the devs and community are skeptical but not altogether opposed if licensing / quality issues are "solved." Kinda milquetoast but I suppose pragmatically an ok choice given the state of things. elementaryOS and Mint are other nice options. But upstream seems like it will leave me with flexibility. I still like the idea of Chimera, both for its overall technical approach and the community around it, but at this point it would probably still require a lot of handholding.

Gaaah! Buffering! #monsterdon

@madargon@is-a.cat @davidgerard@circumstances.run I've been seeing an upsurge in this kind of discourse recently. Generally more discussion around "please dont be mean to the token generators" like this from Troy Hunt a couple of weeks ago https://www.troyhunt.com/weekly-update-501/

@bluebells@tldr.nettime.org @tante@tldr.nettime.org Check out https://openalex.org/ too. The interface is a bit clunky but it's the major open source of scholarly data. (I used to recommend it unequivocally, but unfortunately their CEO and dev are all in on vibe coding so I have less confidence in their data. Still at least as good as the alternatives, though)

@bit101 @sue Absolutely. Even students who don't want to use genai (probably the majority) feel so much pressure in the current higher-ed context of zero-sum grading and multiple conflicting classes.

@sue Having seen the way students I work with use genai, I have to strongly disagree with the author that the tech "is a tremendously effective TA." Even if it could simulate the kind of language TAs use well (questionable), it can't be responsive to the student and the course context in the same way. It will emit a bunch of text, including a maybe-working answer that students copy-paste without really considering it. Genai can't give the kind of context-aware accountability of an instructor.

@steve This bot is pretty good @BCBYNCSA

@pirategoddess@social.vivaldi.net It's genuinely sad. I rely on this data for my research and I'm probably just going to start researching different things now. Another casualty of slop wars.

@kattekrab@aus.social i can narrow down which timeline i ended up in after the Catastrophe

@zeldman@front-end.social i keep going back and rereading this short piece on why generative ai is essentially fascist. https://blog.bgcarlisle.com/2025/05/16/a-plausible-scalable-and-slightly-wrong-black-box-why-large-language-models-are-a-fascist-technology-that-cannot-be-redeemed/

@tuomov@mastodon.social @bluebells@tldr.nettime.org @tante@tldr.nettime.org Ha. Yeah author disambiguation on openalex still has some issues. It's actually one of the things that feels like it has gotten worse as the openalex team has relied more and more on slopcoding