this post was submitted on 31 Oct 2026
73 points (98.7% liked)

technology

24515 readers
317 users here now

On the road to fully automated luxury gay space communism.

Spreading Linux propaganda since 2020

Rules:

founded 6 years ago
MODERATORS
you are viewing a single comment's thread
view the rest of the comments
[–] senseamidmadness@lemmygrad.ml 17 points 1 day ago (2 children)

I think there's a pretty big difference in Tuta's case between "going full AI slop" and "allowing developers to use LLM's for internal coding while still being responsible for its mistakes".

It's a tool being pushed by horrible evil capitalists, but there are ways to use it that aren't stealing from artists and removing jobs. Open-source Chinese LLM's are reportedly very helpful for saving time while programming.

Proton's CEO is delusional though. Someone needs to remove him from that position.

[–] hello_hello@hexbear.net 16 points 1 day ago (1 children)

Chatbot, America stalin-nyet

Chatbot, China stalin-comical-spoon

Never change lemmygrad.

[–] senseamidmadness@lemmygrad.ml 16 points 1 day ago (1 children)

I am tempted to commission or draw a scene of President Xi stealing OpenAI's ice cream with a comically large spoon.

[–] EmmaGoldman@hexbear.net 7 points 1 day ago (2 children)

Oh, you're not gonna proompt it? Interesting.

[–] Tophattem_dawg@hexbear.net 5 points 1 day ago (1 children)
[–] EmmaGoldman@hexbear.net 6 points 1 day ago (1 children)

Just sick of the slop worshippers on the 'grad.

[–] senseamidmadness@lemmygrad.ml 3 points 1 day ago (1 children)

There are slop worshippers on my instance? Every post I see involving GenAI (including prompted memes) gets downvoted to hell. Maybe I am not in the subcommunities where they get accepted

[–] EmmaGoldman@hexbear.net 6 points 1 day ago (1 children)

I see Yogthos shilling this crap on the daily, unfortunately.

How unfortunate. That user's comment history is...very Redditor. I always hope folks on Lemmy are better but some prove me wrong

[–] senseamidmadness@lemmygrad.ml 3 points 1 day ago* (last edited 1 day ago) (1 children)

I don't use generative AI personally, and especially hate all of the Silicon Valley models because the people in charge of them are cartoonishly evil. I've never asked Chat Jibbity or Crock or Clog or Moose or Germ-In-Eye or whatever they're called for anything and never will. My former opinion on LLM's being totally useless was recently changed because another user in my instance told me how helpful DeepSeek has been in saving tedium for open-source software engineering.

[–] EmmaGoldman@hexbear.net 7 points 1 day ago (1 children)

Unfortunately it seems like the amount of time and effort it's saving for open source projects is going to end up causing more problems in the long run.

[–] senseamidmadness@lemmygrad.ml 6 points 1 day ago* (last edited 1 day ago)

What I've heard from developers I know is that Chat Jibbity and Clog Code produce godawful spaghetti code, so they probably are doing a bundle of harm to open-source projects.

[–] Cleisthenian@lemmy.ml 3 points 1 day ago

David Gerard is a rare case of anti ai psychosis imo. The nuances like you mention above are lost on him, and he regularly attacks established foss projects that make the world a better place while he just writes yet another angry blog on the internet. I fully expect him to say ai corrupts the soul and is an affront to god within a year or two.