https://itsfoss.com/news/linux-ai-coding-assistants-policy/anons, what's your stance on this? will you face the ai future or flee to parts unknown?
>>108633484
I'm not reading article typed up by feces
>>108633929there are other articles on this very topic, i personally like to use the linux experiment as my source but hes a youtuber and i figured none of you guys like youtubers so i just picked out whichever article i found on this topic, generally speaking im not a news guy, for the most part i care very little
>>108633849there's quite a few open source models, ya know.
>>108633849>>108634058you are both off topic, the discussion is about ai code being allowed in the linux kernel (with certain standards)
>>108634062nobody cares because open source doesn't give a shit about how you make the code it's how it's licensed.it's like asking what's our thoughts on using vim or emacs
>>108634069naive thinking
The article says linux doesn't need written approval from the core developers to be approved, but it's still reviewed.if code is still scrutinized, iterated upon and improved in pr reviews, who cares?
>>108634069licensing is a concern in vibe coding, indeed, but i am not sure how does a proprietary model impact licensing of one's slop anywayanyway, i am asking the most divided tech corner of the internet i have ever came across, there are those who are vehemently pro-ai or vehemently anti-ai, there are those who think using windows is a flex of wisdom (it isn't), hell, i doubt most people here even know how to write a basic "hello world" program in any language, i am even betting some are willing to prove me wrong and are looking up how to make a hello world program as you read this reply. point is, people have their prejudices, especially here in the chan, this is why i made this thread, i want to see rationales of anons in regards to changes that will no doubt stir controversy
>>108634127licensing is a problem how, you mean "unoriginal code"? a large repo where you're submitting bug fixes is not going to have large weird ai dumps. in real life you're going to tweak code to follow repo conventions, be easy to read and not send a dumb of thousands of lines or many many files without being a senior developer who can defend their code and collaborate. What I'm trying to say is that the (good) solutions are so tweaked, small and spread out that I don't know why it would be a problem. Is that what you meant?
>>108634161ai models are trained on things we have wrote out as mankind earlier, innit? far as i know ai does not seem to care about the licensing of the source material, the training data is shovelled in regardless if its gpl, apache or even bsd.me? i am not thrilled with the idea of generated slop, no matter how hard you polish it, being implemented in such a vital project, i am rather purist in regards of coding.
>>108633484we need a rust-less and AI-less fork of the kernel
>>108633484It's logical to think that Linux will improve its kernel through VibeGODDING.It's simply the superior way of coding., Face it, coders won’t longer be required, and Linux users won't have to wait for a codetrans to implement certain functions. Instead, they'll be able to instruct an AI agent to configure their Linux distribution according to their specifications.Don't like it? Fork it and create your own, luddite (snailcat). AI is here to say. We won
>>108634982This.But isnt Hurd or Linux-libre kind of already that?
>>108635139>hurdnot binary compatible with linux>linux-librecontains rust
>>108635211Why would it need to be binary compatible with Linux?
>>108635219so you don't need a translation layer/emulator to use linux apps?I'm not a gentroon to compile everything myself
>>108634982Arent the rust modules optional anyway?I dont see why anyone would care about that though
redhat and ibm sell ai shit this was bound to happen eventually
>>108635289>I dont see why anyone would care about that thoughbecause it pulls in dependencies during compile time, meaning the whole language is an attack vector
>>108635328Is make an attack vector?
>>108635350you can easily control and *verify* everything make does the same doesn't apply to rust
>>108635372You can do that in cargo too (or virtually any language other than C that does the same thing)
>>108634262polishing? well, I'm talking about bug fixes and even some features. You're not gonna come up with anything new even if you code those by hand. We're talking about fixing already existing patterns in code and implementing new things based on utils and shared code in the repo itself. Again, we're not talking about a vibecoder asking 'generate linux 2 make no mistakes' but the huge backlog of reported bugs nobody has time to deal with. And if we don't even want to touch generation, we also just have help debugging, which AI also helps with when the logic and parts of something is scattered across many files
https://github.com/openbsd/src/commit/9c2b8e445a0bdfafdd6148b1760f00aa5429627bhttps://github.com/openbsd/src/commit/e9af5eb5a61d189327b553b24d0d31f19c64b63f
Linus has been such a complete joke since they sent him to diversity camp.He may as well die and trigger the succession now he's just their meat puppet anyway.
>>108635084what happens when this guy realizes companies and oss projects are still going to employ/maintain actual SWE’s to do the ai coding
>>108635084
>>108635084Not to sound racist, but this post smells like curry...
>>108635542Cope, you will be replaced>>108636360Gay>>108636380>Indians out of nowhere
>>108635244Well that sounds like a you problem then.
>>108636380Stop replying to curry bots.
>>108634109code review can only catch so much, the main problem I see is the volume of code AI can produce, until now all our software production has been bound by how much a human can work. If more code starts coming in, is it really going to get properly scrutinized
