KhrysMar 13
RT if you want a CLEAR statement about AI from all GNU/Linux distributions and are ready to quit any distribution that is ok with integrating AI slopware.
Show threadFlorianMar 13

@Khrys Ready to quit ?
It depends ... what is the degree of cleaness that we consider ?

I may be wrong, but isn't the linux kernel project already accepting contributions made with the use of AI-generated code ?

If this is the case, then there is no such GNU/Linux AI-free distribution anymore.

Edit: to avoid confusion, that would be something I regret. But would I have the strength to resist it ? I don't know.

Show threadmilagemayvaryMar 13

@fmasy @Khrys

AFAIK GNU doesn't, got a source?

Show threadtechnomancyMar 13

@milagemayvary @fmasy @Khrys GNU is a group of loosely-associated projects

they absolutely can and should have a clear stance on this topic, but for whatever reason they have failed to publish a clear statement yet

(this is all the more baffling because they have decades of experience taking hardline moral positions against strong community objection, so why would they be dragging their feet on this extremely urgent one???)

until that happens, I'm hesitant to assume that any project will be clean just because it's under the GNU umbrella

Show threadlhp
@technomancy @milagemayvary @fmasy @Khrys have they actually taken any new hardline stances in the last two decades? They might have just lost their bite. Which is unfortunate; if emacs falls to slop I will be mad.
Mar 13 at 5:22pmMastodon for Android
Show threadtechnomancyMar 13

@lhp @milagemayvary @fmasy @Khrys yeah, the truth is probably that they are just so myopically focused on licensing and DRM that they can't acknowledge the importance of any other concern

huge amount of wasted potential; it's especially sad in this case where we urgently need leadership and coordination

Show threadPunnO)))Mar 13

@technomancy @lhp @milagemayvary @fmasy @Khrys Which makes no sense to me, LLMs are the greatest threat to software licensing in our lifetimes.

...so far

Show threadtechnomancyMar 13

@ieure @lhp @milagemayvary @fmasy @Khrys another angle here that I'm trying to track down is that if the contributor agreement requires you to be submitting an original work that you hold copyright to, then they might already ban LLM-generated patches by default

however the FSF bizarrely does not publish their copyright assignment paperwork as far as I can tell and goes to some lengths to prevent it from being posted publicly

Show threadTorbjörn NorinderMar 13
@technomancy @ieure @lhp @milagemayvary @fmasy @Khrys there is a clause requiring the signee to warrant being the sole copyright holder, at least in the papers I signed in 2024
Show threadtechnomancyMar 13

@tnorinder @ieure @lhp @milagemayvary @fmasy @Khrys I think some maintainers might argue that LLM-generated code may bypass the paperwork altogether as it's not under copyright so there's nothing to assign

in any case, it's urgently needed for GNU leadership to clarify a policy, but I don't have a lot of hope they will

Show threadAnthony SoraceMar 13

@lhp @technomancy @milagemayvary @fmasy @Khrys

"have they actually taken any new hardline stances in the last two decades?"

Does "we're okay keeping serial harassers in positions of power" count as a hardline stance? Feels like a "moral stance" in one direction, anyway.

Given RMS's own history with the source for Emacs, I wouldn't look to them for moral leadership resisting the fascist plagiarism machine.