Don't use Fedora Linux https://mastodon.social/@gamingonlinux/115422841053394776
-
-
-
@sauc3 @mcc @slembcke if this is the news I think we're talking about, the Fedora project leadership voted to allow "ai" generated contributions provided that substantial uses of it are labeled accordingly. This would mean that software that is developed and maintained by the Fedora project may have "ai" generated code in it soon, but that does not necessarily mean that other Linux distributions are "infected". It also does not necessarily mean that other Linux distributions are not "infected"
-
@sauc3 @mcc @slembcke if this is the news I think we're talking about, the Fedora project leadership voted to allow "ai" generated contributions provided that substantial uses of it are labeled accordingly. This would mean that software that is developed and maintained by the Fedora project may have "ai" generated code in it soon, but that does not necessarily mean that other Linux distributions are "infected". It also does not necessarily mean that other Linux distributions are not "infected"
-
-
@sauc3 @mcc @slembcke also, the more contributors a project has the harder it is to guarantee none of them tried to pass off "ai" generated code as their own and managed to get it into review. Having a policy to at least require it be labeled so it can be given appropriate scrutiny might result in less of it getting in than an outright ban (and definitely less than having no policy at all). Hard to say, I'd rather ban it, but people are more honest when the answer is "maybe" instead of "no".
-
@sauc3 @mcc @slembcke also, the more contributors a project has the harder it is to guarantee none of them tried to pass off "ai" generated code as their own and managed to get it into review. Having a policy to at least require it be labeled so it can be given appropriate scrutiny might result in less of it getting in than an outright ban (and definitely less than having no policy at all). Hard to say, I'd rather ban it, but people are more honest when the answer is "maybe" instead of "no".
-
-
@aeva @sauc3 @mcc @slembcke the Fedora position on AI is driven by RedHat putting AI coding up as a KPI for the devs to get a bonus, so you can see it as the CEO mandate you mention. And sadly, RedHat controls the development of a lot of FLOSS that has been more or less shine e down everyone's throat across all distributions. 8-(
-
@sauc3 @mcc @slembcke also, the more contributors a project has the harder it is to guarantee none of them tried to pass off "ai" generated code as their own and managed to get it into review. Having a policy to at least require it be labeled so it can be given appropriate scrutiny might result in less of it getting in than an outright ban (and definitely less than having no policy at all). Hard to say, I'd rather ban it, but people are more honest when the answer is "maybe" instead of "no".
@aeva @sauc3 @mcc @slembcke fwiw the reason mesa added a policy about it is just that One Guy started creating totally pointless mega MRs with nothing of value in them, and they had no existing rule to say "don't do that". There's been no actual policy really proposed to either ban or embrace it yet.