1. Debian Developer resigns from the project in response to the recent vote to allow generative AI use within Debian.2.
-
-
@jik good on you
-
Well, the author of the abusive message above on the left was banned from Debian's mailing lists less than two hours after he sent it, and three people (including me) have objected to what he said, and no one (yet) has tried to justify or support it, so that's something anyway.
@jik I'm starting to realize AI is a cult. Its defenders sound more and more like the unhinged Tulsi stans who infested Twitter back in the day.
-
@jik I'm starting to realize AI is a cult. Its defenders sound more and more like the unhinged Tulsi stans who infested Twitter back in the day.
@msbellows @jik Someone said "Everyone for AI is too for it, and everyone against AI is too against it" and that nails it. Ideologues are noisy at both ends of the spectrum.
-
To ai sloppers: ai slop is literally destroying everything, now you want to destroy Debian with it. You ai bros are morons, and it will be nice when you're all forgotten.
-
To ai sloppers: ai slop is literally destroying everything, now you want to destroy Debian with it. You ai bros are morons, and it will be nice when you're all forgotten.
@KittenKoder hey, if you're coming after me as an AI bro then you've misunderstood the situation, and if you're not, then you might want to be more precise in your language in the future
-
@KittenKoder hey, if you're coming after me as an AI bro then you've misunderstood the situation, and if you're not, then you might want to be more precise in your language in the future
@jik I misread it, so my post will be to the ai slopper.
-
@msbellows @jik Someone said "Everyone for AI is too for it, and everyone against AI is too against it" and that nails it. Ideologues are noisy at both ends of the spectrum.
@mogul @msbellows @jik
No. That's a generalization too far. A representative majority of both 'sides' is not the same. The people who boost AI in business can be independently shown to be further from measurable reality than those who don't. Proper, peer reviewed research has shown traits correlate with liking AI. "The most consistent predictors of AI use across studies were aversive personality traits (e.g., Machiavellianism, narcissism, psychopathy)" https://pubmed.ncbi.nlm.nih.gov/40977447/ -
-
@mogul @msbellows @jik
No. That's a generalization too far. A representative majority of both 'sides' is not the same. The people who boost AI in business can be independently shown to be further from measurable reality than those who don't. Proper, peer reviewed research has shown traits correlate with liking AI. "The most consistent predictors of AI use across studies were aversive personality traits (e.g., Machiavellianism, narcissism, psychopathy)" https://pubmed.ncbi.nlm.nih.gov/40977447/@AlexsandraSmart @msbellows @jik I'm not doubting that. But I'm also seeing people opposed to AI apply use and/or tolerance of AI as purity tests for projects, people, and purpose, and rejecting any association with anyone who fails their test because of the people off the deep end you're talking about.
-
@jik tone of the message aside. How is #Debian supposed to ban AI assisted coding or only include human generated #code?
I'm not defending what or how the guy said what he said, it was terrible antisocial nerd cringe.
But shouldn't we be focusing on the quality of the code submissions? How much can AI touch code I submitted? If I use it to keep my TODO list or as a rubber duck should my code be excluded?
I'm asking in good faith, not trolling.
-
@jik tone of the message aside. How is #Debian supposed to ban AI assisted coding or only include human generated #code?
I'm not defending what or how the guy said what he said, it was terrible antisocial nerd cringe.
But shouldn't we be focusing on the quality of the code submissions? How much can AI touch code I submitted? If I use it to keep my TODO list or as a rubber duck should my code be excluded?
I'm asking in good faith, not trolling.
@MeaningfulBits @jik The way a ban on AI-assisted coding would work is that people agree not to do it as a condition of contributing to the project and are expected to keep their promise, and violating that agreement would be a violation of the code of conduct.
It’s not really very complicated and I don’t know why AI lovers want to make it more complicated than it is.
-
@AlexsandraSmart @msbellows @jik I'm not doubting that. But I'm also seeing people opposed to AI apply use and/or tolerance of AI as purity tests for projects, people, and purpose, and rejecting any association with anyone who fails their test because of the people off the deep end you're talking about.
@mogul @AlexsandraSmart @jik Except AI isn't the baby thrown out with the bathwater (its more rabid advocates). AI is objectively harmful. It is trained on stolen IP, wastes massive amounts of water, requires massive energy inputs when climate change demands that we reduce energy use, is making the open Internet almost unusable, and measurably makes users dumber. I'll agree it has limited usefulness for certain kinds of scientific research and for disability accommodations; other than that, it's an evil in and of itself.
-
@jik This is so depressing. Debian is the Linux distro I’ve put my trust on over the last 25+ years. I can’t even process this.
@fabi1cazenave @jik switch to #runbsd
-
@mogul @AlexsandraSmart @jik Except AI isn't the baby thrown out with the bathwater (its more rabid advocates). AI is objectively harmful. It is trained on stolen IP, wastes massive amounts of water, requires massive energy inputs when climate change demands that we reduce energy use, is making the open Internet almost unusable, and measurably makes users dumber. I'll agree it has limited usefulness for certain kinds of scientific research and for disability accommodations; other than that, it's an evil in and of itself.
@msbellows @AlexsandraSmart @jik All fair criticisms, and ones that I hope will apply less in time with local inference, ethically-sourced models, etc. The technology is definitely in a "fetal stem cell" period of its existence.
-
@jik tone of the message aside. How is #Debian supposed to ban AI assisted coding or only include human generated #code?
I'm not defending what or how the guy said what he said, it was terrible antisocial nerd cringe.
But shouldn't we be focusing on the quality of the code submissions? How much can AI touch code I submitted? If I use it to keep my TODO list or as a rubber duck should my code be excluded?
I'm asking in good faith, not trolling.
Just as copyright violation in #Debian is impossible to perfectly detect, and yet is strictly banned in the project, @MeaningfulBits.
It doesn't matter whether a perfect detection mechanism exists (it doesn't). The point of a ban is to be clear *that* this is the policy, *what* behaviour is banned, and that people know it.
Yes people can attempt to deliberately circumvent it. So we don't want people who deliberately try to evade policy.
-
@jik tone of the message aside. How is #Debian supposed to ban AI assisted coding or only include human generated #code?
I'm not defending what or how the guy said what he said, it was terrible antisocial nerd cringe.
But shouldn't we be focusing on the quality of the code submissions? How much can AI touch code I submitted? If I use it to keep my TODO list or as a rubber duck should my code be excluded?
I'm asking in good faith, not trolling.
@MeaningfulBits
> But shouldn't we be focusing on the quality of the code submissions?“Leaving aside the ethical concerns, shouldn't we be focusing on the productive work that child labour can achieve?”
“I'm not defending the practice of stuffing asbestos into an apartment building's walls. But shouldn't we be focusing on the quality of insulation?”
There is no ethical way to dismiss the ethical concerns. Until they are properly addressed, no we should not focus on quality of the code.
-
@MeaningfulBits @jik The way a ban on AI-assisted coding would work is that people agree not to do it as a condition of contributing to the project and are expected to keep their promise, and violating that agreement would be a violation of the code of conduct.
It’s not really very complicated and I don’t know why AI lovers want to make it more complicated than it is.
@benjamineskola @jik who determines which is which. Just seems like it would be hard to enforce at best and miss used at worse. For a project that already has limited resources do we hire an #AI code watcher?
-
@MeaningfulBits
> But shouldn't we be focusing on the quality of the code submissions?“Leaving aside the ethical concerns, shouldn't we be focusing on the productive work that child labour can achieve?”
“I'm not defending the practice of stuffing asbestos into an apartment building's walls. But shouldn't we be focusing on the quality of insulation?”
There is no ethical way to dismiss the ethical concerns. Until they are properly addressed, no we should not focus on quality of the code.
@bignose @jik false equivalency aside. I rather label the person submitting the code responsible for w/e fuck ups come out of it instead of blaming the tool.
A local hobby model crossing off my TODO list and rubber ducking doesn't fit your narrative.
Yes children making shoes are bad, should I stop wearing shoes or jail the ceo?
Yes cancer causing material shouldn't be in our homes should I stop using insulation or jail the builder who put bad product in homes.
-
@benjamineskola @jik who determines which is which. Just seems like it would be hard to enforce at best and miss used at worse. For a project that already has limited resources do we hire an #AI code watcher?
@MeaningfulBits @jik Why would you need to enforce it? People would have to agree to do it as a condition of membership in the project. I’m sure they can be trusted to keep their promise — or are you saying all AI users are liars who need constant monitoring to ensure they’re following the rules?
