Linux kernel goes full AI slop, hides chatbot commits
YouTube
Discussion
YouTube
@jupiter But Torvalds putting out far right dogwhistles is deeply painful:
You're right that he could have made his point without repeating far-right talking points.
He could have said (as he has many times before) something like "This project's goal is practical, not philosophical and I reject calls to reject a technology on a philosophical basis"
One might agree or disagree, but that would have said the same thing without the problematic language.
@mpjgregoire for self-labelling, I prefer using precise labels.
Because with overly generalized labels, you imply that doing the right thing means the same for everyone.
I realized around 15 years ago how much this is not the case – and turned the core of this understanding into a roleplaying supplement:
https://www.1w6.org/english/core-drivers
@serge @jupiter
@jupiter And I wonder whether we should have expected that, because doing something for idealism vs. doing something for technology and not trusting idealists has always been a conflict between Linus and GNU – with Linus not even accepting GPLv2 *or later*, because he did not trust the idealists from the FSF.
Which is the reason why the EU had to rule that getting updates to smartphones is required for being in the EU market–while GPLv3 would have guaranteed by itself that you can update.
@ArneBab Hmm. I'm still waiting for HURD to be honest, so GNU has become quite the second violin to Linux, rightfully or not. I do think the AGPL3 is one of the best and purest things to license under, though.
@jupiter there’s a bit more nuance in the argument, but it shows that once you let AI coders in, some will try to push you out – an excerpt from LWN (only an excerpt because it’s subscriber only content):
https://rollenspiel.social/@ArneBab/117447227490312321
https://lwn.net/Articles/1099594/